Chapter illustrations ()
Loading image navigation...
No image index could be found for this publication.
Close this Panel
2
The Geographical Tradition
Medieval Europe thus inherited the geographical ideas of the classical world, but in a calcified and increasingly mythologized form. Whereas the best Greek geographers had recognized the conventional nature of the continents – and insisted that the Red Sea made a more appropriate boundary between Asia and Africa than the Nile River – such niceties were often lost on their counterparts in late antiquity and the early Middle Ages. … But until the late Middle Ages, reference to the larger formal continental scheme was largely limited to recondite geographical studies, finding little place in general scholarly discourse.1Martin W. Lewis and Kären E. Wigen, The Myth of Continents: A Critique of Metageography (Berkeley: University of California Press, 1997), 23–4; for an accurate and nuanced treatment of this issue, see Mauntel, Erdteile.
Although historians of medieval geography have long recognised that the study of geography in the Middle Ages was hardly stagnant, there remains a pervasive notion in the wider scholarly community that, at least in terms of its formal structures, medieval geography was a fundamentally passive receptor of the ancient geographical tradition.2On medieval geographical innovations, see already Raymond Beazley, The Dawn of Modern Geography, 3 vols (London: John Murray, 1897–1906), 2:8–13. The reason for this is not hard to see, since medieval authors themselves were often at pains to stress their deference to older authorities. The late ancient formula for the division of the world into Asia, Europa and Africa, for example, remained ubiquitous and largely unquestioned well into the sixteenth century.3Mauntel, Erdteile, 77–8, 282–5, 580, passim. No less emblematic of this deferential attitude is Otto of Freising’s offhanded suggestion that those who wish to know more about Asia, Europe and Africa should simply consult Orosius – an approach that is still alive and well in the 1539 geography of Jacques Signot.4Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 38, trans. Charles C. Mierow (New York: Columbia University Press, 2002), 123: ‘Quarum provintias, situs, regiones qui velit cognoscere, legat Orosium’ (If anyone wishes to know the provinces, the topography and the divisions of these parts of the world, let him read Orosius). Cf. Hon., Im. mun. prol. Christ., ed. Flint, 49: ‘Nichil autem in eo pono nisi quod majorum commendat traditio’ (I have set down nothing in this book except what the tradition of our ancestors has commended). On Signot’s La division du monde, see Mauntel, Erdteile, 580. So also the fact that even where authors depart from the late ancient tradition, there is an overwhelming tendency towards harmonisation or the identification of novelties with (often unrelated) Roman provinces or toponyms.5Felicitas Schmieder, ‘Gogs und Magogs “natürliche Milde”? Die Mongolen als Endzeitvölker im Wandel von Wissen und Wünschen’, in Peoples of the Apocalypse: Eschatological Beliefs and Political Scenarios, ed. Wolfram Brandes, Felicitas Schmieder and Rebekka Voß (Berlin: De Gruyter, 2016), 112, and compare Schmieder’s treatment of Gog and Magog with Remi of Auxerre’s letter about the Hungarians in R. B. C. Huygens, ‘Un témoin de la crainte de l’an 1000: la lettre sur les Hongrois’, Latomus 15, no. 2 (1956): 225–39, esp. ll. 96–101, 161–201; cf. Claude Nicolet and Patrick Gautier Dalché, ‘Les “quatre sages” de Jules César et la “mesure du monde” selon Julius Honorius: réalité antique et tradition médiévale’, Journal des savants 4, no. 1 (1986): 197–8n53. See also Wolever, ‘Post-Carolingian Geographies’, 153–63. Yet as I will argue in this chapter, this overt conservatism was often but a cover concealing more substantial underlying developments and by remaining attentive to often subtle changes we can identify consistent critical engagement with those same late antique authorities and widespread structural developments in the way that medieval authors conceptualised and engaged with the world.
To understand why the formal geographical literature of the twelfth century can appear so repetitive and backward-looking, it is important to understand the enigmatic position that geography held within the broader intellectual environment. It remained what might be termed an auxiliary science, which was rarely studied or taught on its own terms, but rather in aid of another subject, such as to learn biblical toponyms or as a prelude to a historical work.6Natalia Lozovsky, ‘The Earth Is Our Book’: Geographical Knowledge in the Latin West ca. 400–1000 (Ann Arbor: University of Michigan Press, 2000), ch. 1; Deluz, ‘Une image du monde’, 44. Despite this auxiliary status and the continued dominance of a small canon of late antique geographies, the early decades of the twelfth century saw an unusual proliferation of new and influential geographical compilations and descriptions of the inhabited world. These texts were driven centrally by the expansion of education over the eleventh and particularly the early twelfth centuries.7Deluz, ‘Une image’, 41. On the expansion of education in the period, see Ferroulo, Origins of the University. On the expansion of monasteries, see Giles Constable, The Reformation of the Twelfth Century (Cambridge: Cambridge University Press, 1996), 88–93. Indeed, important geographical texts by Honorius Augustodunensis and Hugh of Saint Victor are explicitly educational in aim.8Hon., Im. mun. prol. Christ., ed. Flint, 48–9; HoSV, Desc. prol., 12, ed. Gautier Dalché, 133; Patrick Gautier Dalché, ‘Nouvelles lumières sur la Descriptio mappe mundi de Hugues de Saint-Victor’, in Géographie et culture: La représentation de l’espace du VIe Au XIIe siècle (Farnham: Ashgate, 1997), 1–27; cf. William M. Green, ‘Hugo of St Victor: De Tribus Maximis Circumstantiis Gestorum’, Speculum 18, no. 4 (1943): 484–93. On Guido of Pisa, see Campopiano, ‘Construction of the Text’, 82. Importantly, however, they coincide with a moment of significant expansion of geographical knowledge, which these authors increasingly sought to integrate into this new era of geographical textbooks. As a result, and due in no small part to this educational context, these twelfth-century geographies represent a significant moment of consolidation for the Latin geographical tradition. This is evident not only in influential written geographies like Honorius’s Imago mundi, but also in the proliferation of maps and in particular the solidification of an influential style of circular, ‘Anglo-French’ mappa mundi, which appears to have spread especially from the schools of northern France and even Hugh’s own classroom.9P. D. A. Harvey, ‘The Sawley Map and Other World Maps in Twelfth-Century England’, Imago Mundi 49, no. 1 (1997): 37–8; Dan Terkla, ‘Hugh of St Victor (1096–1141) and Anglo-French Cartography’, Imago Mundi 65, no. 2 (2013): 161–79; Dan Terkla, ‘Introduction: Where to fix Cadiz?’, and Nathalie Bouloux, ‘The Munich Map (c. 1130): Description, Meanings and Uses’, in A Critical Companion to English Mappae Mundi of the Twelfth and Thirteenth Centuries, ed. Dan Terkla and Nick Millea (Woodbridge: Boydell, 2019), 10–11, 93.
It was in these schoolrooms of northern France that the geographical output of the twelfth century saw increasing discontinuity with the authors of late antiquity. Here we find authors not only updating their geographies for a modern audience, but actively reshaping them both structurally and materially. There are two sets of interconnected issues that are of interest. First, regarding the way in which medieval authors divided the world into different regions, I will show how these regions are not simple, transhistorical categories, but instead developed out of a complex historical genealogy and remained contested categories into the High Middle Ages. Second, regarding the structural and organisational principles that underlie the way that medieval geographical texts were both produced and read, I will address how medieval texts structured geographical space on a practical level, and will again highlight how, despite its frequent conservative pretence, this geographical tradition was neither static nor homogeneous. Finally, the chapter will conclude with a brief reflection on how this geographical tradition relates back to the exegetical tradition and what that says about the place of the cardinal points as cosmological categories. Across these sections I will show how, by the twelfth century, the cardinal directions had come to play a fundamental role in dividing and structuring world geography as well as situating it cosmologically.
Dividing the World
The division of the world into three parts – Asia, Europe and Africa – has an extensive history stretching back to Greek antiquity. Though the medieval world up to the twelfth century had little direct access to Greek texts, there were at least two channels through which Ancient Greek geography still directly impacted the Latin Middle Ages: through late antique references and through later translations, mostly via Arabic.10Lozovsky, Earth Is Our Book, 6–10; Patrick Gautier Dalché, La géographie de Ptolémée en Occident (IVe–XVIe siècle) (Turnhout: Brepols, 2009). More fundamentally, medieval authors stood within the lineage of the Greek geographical tradition as mediated by and developed through the Roman world. Although there is a tendency to imagine that this ancient geographical tradition provided a largely clear and uncomplicated understanding of the inhabited world, the reality is more complex. Classical geography left a wealth of ambiguities and contradictions that continued to structure the geographical tradition across the Middle Ages.11For example, see the confusion over Thule, Balcia and Scandinavia (Fraesdorff, Norden, 46–50). In order to unravel some of these modern assumptions and frame the problems of geographical division that medieval authors faced, a brief survey of the ways in which the known world was divided down to late antiquity is in order, focusing on three groups of texts: pagan geography down to the first century CE, Christian geography among the Fathers and late antique textbooks.
From Greece to Late Antiquity
The general trends for dividing the oecumene – that is, the inhabited world, generally understood as the combined Afro-Eurasian landmass – which would hold for the next two thousand years, were more or less established by the fifth century BCE. This story begins with the tripartite division of the oecumene by Ionian geographers in the late sixth to early fifth centuries BCE and the geographic dichotomies between north-south and east-west that accompanied it. The result was a series of two- and three-part models. Europe and Asia, the two primary parts, were either divided north-south by the river Phasis, thought to run directly west from the Black Sea, or east-west by the river Tanais, the standard division from the Hellenistic era.12James Romm, ‘Continents, Climates, and Cultures: Greek Theories of Global Structure’, in Geography and Ethnography: Perceptions of the World in Pre-Modern Societies, ed. Kurt A. Raaflaub and Richard J. A. Talbert (Chichester: Wiley-Blackwell, 2010), 215–20. The Tanais and Phasis correspond to the modern rivers Don and Rioni. Herodotus describes these conflicting conventions in his criticism of such divisions, noting first his amusement at those who represent the earth with a perfect circle ‘as if drawn with a pair of compasses’ divided equally into Asia and Europe. He likewise expresses his amazement that the singular world should be given three names (Libya, Asia and Europe) and that the Nile and the Phasis, or Tanais as some prefer, should divide them.13Herodotus, Histories 4.35, 42, trans. Robin Waterfield (Oxford: Oxford University Press, 1998), 247–8; see Rosalind Thomas, Herodotus in Context (Cambridge: Cambridge University Press, 2002), ch. 3, and Reinhold Bichler, ‘Der Antagonismus von Asien und Europa – eine historiographische Konzeption aus Kleinasien?’, in Der Beitrag Kleinasiens zur Kultur- und Geistesgeschichte der griechisch-römischen Antike, ed. Josef Fischer (Vienna: Austrian Academy of Sciences, 2014), 10–11. In the end, rather than resolving this dilemma, Herodotus simply follows the standard tripartition.14Germaine Aujac, ‘The Foundations of Theoretical Cartography in Archaic and Classical Greece’, in The History of Cartography, vol. 1, ed. J. B. Harley and David Woodward (Chicago: University of Chicago Press, 1987), 135–7. Despite his scepticism towards these rigid divisions of the world, elements of a north-south, hot-cold climatic theory litter the Histories more broadly, though they remain unsystematic, and perhaps, as Isaac suggests, simply a cultural commonplace.15Romm, ‘Continents, Climates, and Cultures’, 218–20; Isaac, Invention of Racism, 56–60. For a more extensive discussion of the ways in which a north-south dichotomy takes precedence over east-west, see James Romm, ‘Herodotus and Mythic Geography: The Case of the Hyperboreans’, Transactions of the American Philological Association 119 (1989): 97–113.
While Herodotus does not present an east-west opposition between Europe and Asia, subsequent generations do take steps in this direction.16Isaac, Invention of Racism, ch. 4; cf. Bichler, ‘Der Antagonismus’, 12–14. The most important text in this regard is the Hippocratic Airs, Waters, Places of the late fifth century BCE, likely from the generation after Herodotus.17Isaac, Invention of Racism, 60n18; Thomas, Herodotus, 86. The Airs makes a number of developments that will prove crucial for subsequent conceptions of the division of the world. In the first place, the Airs shifts the focus decidedly towards a dichotomy between Europe and Asia. Chapter twelve confidently declares that it will ‘show how Asia and Europe differ from one another in every respect’.18Airs, Waters, Places 12, trans. Paul Potter, Hippocrates, vol. 1, Loeb Classical Library 147 (Cambridge, MA: Harvard University Press, 2022), 109; Isaac, Invention of Racism, 61. As a result of major lacunae in the text, it is unclear whether the author treated Libya as an independent region or as a subdivision of Asia; nevertheless, the professed focus on Europe and Asia drives the remainder of the extant text.19Romm, ‘Continents, Climates, and Cultures’, 220–3, see 232n21 for discussion on whether Libya is treated as a continent in its own right. Secondly, the author follows the east-west division of Europe and Asia, noting the Meotide swamps as the boundary.20Airs, Waters, Places 12–13, trans. Potter, 109–15. Lastly, unlike Herodotus, the Airs is systematic in its application of environmental determinism. The asymmetrical impact of the east and west wind governs the environment, and thus peoples, of Europe and Asia. The Airs is, however, unusual in its development of an intricate model of east and west alongside north and south as environmental categories – most subsequent authors preferred the more straightforward mapping of hot and cold onto north and south.21Romm, ‘Continents, Climates, and Cultures’, 221–3.
The last major development came in the fourth century BCE with the luminaries of classical antiquity. In Greek literature, a strong opposition between Europe and Asia only emerged in the generations after the Greco-Persian war, and perhaps in response to the Peloponnesian war, in the context of antipathy towards Persia.22Isaac, Invention of Racism, 283–5; Romm, ‘Continents, Climates, and Cultures’, 223–6; cf. Bichler, ‘Der Antagonismus’, 14–16. We can see this developing from the end of the fifth century, but it flourishes particularly in the early fourth century, especially in the oration of Isocrates, who develops a clear dichotomy between Asia and Europe like in the Airs, but now as a polemic against the Persians.23Romm, ‘Continents, Climates, and Cultures’, 225–6. With this shift came a solidification of the Tanais as the boundary of Europe and Asia within a tripartite structure, as these boundaries cease to be a point of serious discussion.24Romm, ‘Continents, Climates, and Cultures’, 223–4, 226. So also with the north-south climatic system, as Plato, for example, sets a north-south zonal division of the world in Socrates’s mouth in the Republic or more systematically in Aristotle’s discussion of climatic determinism in the Politics.25Plato, Republic 4.435e–436a, trans. G. M. A. Grube, in Plato: Complete Works, ed. John M. Cooper (Indianapolis: Hackett, 1997), 1067; Aristotle, Politics 7.7.1327b, trans. Jonathan Barnes, The Complete Works of Aristotle, 2 vols (Princeton: Princeton University Press, 1984), 2:2107; cf. Isaac, Invention of Racism, 73. Finally, we see increasing Hellenic exceptionalism, with Greece being identified as the median point in the world and an increased emphasis on the distinction between Greek and barbarian.26Isaac, Invention of Racism, 283–98; Romm, ‘Continents, Climates, and Cultures’, 224–6. Yet as much as we might discuss a standard east-west division of Europe and Asia emerging alongside a north-south climatic opposition, the conflation and confusion of these categories continued into the Hellenistic era, as we can see in Strabo’s story about how Alexander the Great misidentified the Tanais as an eastward flowing river.27Romm, ‘Continents, Climates, and Cultures’, 226–7; Strabo, Geography 11.7.4, cf. 11.5.5, trans. Horace Leonard Jones, 8 vols, Loeb Classical Library 49, 50, 182, 196, 211, 223, 241, 267 (Cambridge, MA: Harvard University Press, 1917–32), 5:255, cf. 5:239.
There is a common narrative in the scholarship that, with the transition from Greece to Rome, there was a concurrent shift from east-west geographical opposition to north-south.28E.g., Isaac, Invention of Racism, 83–5; Guy Halsall, Barbarian Migrations and the Roman West 376–568 (Cambridge: Cambridge University Press, 2007), 46–7. Though there is some truth here, the reality is, as we have already seen, rather more complicated in the Greek world, which maintained a strong north-south tradition and focused more on the dichotomy of Europe and Asia than of east and west; the reality is no less complicated with republican and early imperial Roman literature. The Ionian division of the oecumene into north and south, often subsequently attributed to Eratosthenes, proved quite resilient.29E.g., Varro, Res rusticae 1.2.3–4, ed. Georg Goetz (Leipzig: Teubner, 1929), 10: ‘cum orbis terrae divisus sit in duas partes ab Eratosthene maxume secundum naturam, ad meridiem versus et ad septemtriones …’ (Since the earth was most naturally divided into two parts by Eratosthenes, towards the south and toward the north …); Strabo, Geography 2.1.1, trans. Jones, 1:223. It continues to pop up across the Hellenistic era down to the first century BCE, even alongside the standard tripartition of the world. Thus for Polybius, writing in the second century BCE, the world was divided into Asia, Africa and Europe by the Tanais, Nile and Pillars of Hercules, yet Europe still opposed Asia and Africa, running from east to west along the north side of the Mediterranean, which universally divides north from south.30Polybius, Histories 3.37.2–8, trans. W. R. Paton, vol. 2, Loeb Classical Libary 137 (Cambridge, MA: Harvard University Press, 2010), 95–7. Likewise for Varro, Europe and Asia simply divided the world into north and south.31Varro, De lingua latina 5.5.31, ed. Georg Goetz and Fritz Schöll (Leipzig: Teubner, 1910), 11: ‘Ut omnis natura in caelum et terram divisa est, sic caeli regionibus terra in Asiam et Europam. Asia enim iacet ad meridiem et austrum, Europa ad septemtriones et aquilonem’ (As the whole of nature is divided into earth and sky, so the earth is divided by the regions of the sky into Asia and Europe. For Asia lies to the south and south-wind while Europe lies to the north and north-wind). Cf. Varro, Res rusticae 1.2.3–4, ed. Goetz, 10. The persistence of this bipartition can be still be felt in Pliny’s Naturalis historia, which, after noting the standard tripartition, suggests that, given the great merits of Europe, many think the world should be divided into two equal parts by the strait of Gibraltar and the river Tanais.32Pliny, Naturalis historia 3.1.5, ed. and trans. Rackham, 2:4–7: ‘Primum ergo de Europa, altrice victoris omnium gentium populi longeque terrarum pulcherrima, quam plerique merito non tertiam portionem fecere, verum aequam, in duas partes ab amne Tanai ad Gaditanum fretum universo orbe diviso’ (To begin then with Europe, nurse of the race that conquered all the nations, and by far the loveliest portion of the earth, which most authorities, not without reason, have reckoned to be not a third part but a half of the world, dividing the whole circle into two portions by a line drawn from the river Don to the Straits of Gibraltar). Cf. Polybius, Histories 3.37.4, trans. Paton, 95, and Strabo, Geography 2.4.5, 11.7.4, trans. Jones, 1:411–13, 5:255.
There were, however, some important developments in Roman geography that would prove decisive for the patristic and medieval world. Central among these was a ninety-degree shift in the dividing axis of the world’s bipartition, which would ultimately establish Asia as fundamentally eastern, rather than southern. The exact causes of this shift are not entirely clear, but it can be plausibly attributed to the expansion of Asia in Roman geography after the campaigns of Alexander the Great, which ultimately undercut the older notion that Europe would equal Asia and Libya together in size.33Claudia Wick, M. Annaeus Lucanus Bellum civile liber IX Kommentar (Munich: K. G. Saur, 2004), 159–60; Benet Salway, ‘Putting the World in Order: Mapping in Roman Texts’, in Ancient Perspectives: Maps and Their Place in Mesopotamia, Egypt, Greece, and Rome, ed. R. J. A. Talbert (Chicago: University of Chicago Press, 2012), 213–14; cf. Mauntel, Erdteile, 102–3, who remains sceptical about Asia’s geographical priority in the Roman period. Our first source for this new image is Sallust’s Bellum Iugurthinum, which for the first time adjoins Africa to Europe and not Asia, explaining that it is commonly regarded as the third part of the world, but that some only recognise Asia and Europe, giving Africa in the latter.34Sallust, Bellum Iugurthinum 17.3, ed. Alfons Kurfess (Leipzig: Teubner, 1957), 68: ‘In divisione orbis terrae plerique in parte tertia Africam posuere, pauci tantummodo Asiam et Europam esse, sed Africam in Europa’ (In the division of the world, the majority consider Africa as a third part, while a few maintain there is only Asia and Europe, but set Africa within Europe). This view quickly became the standard by the early empire; both Pomponius Mela and Strabo also emphasise the great size of Asia in comparison to both Europe and Africa.35Pomponius Mela, De chorographia 1.9, ed. Alain Silberman (Paris: Les Belles Lettres, 1988), 3–4; Strabo, Geography 17.3.1, trans. Jones, 8:155. In the latter, Strabo seems to imply a potential combination of Europe and Africa when he suggests that even combined with Europe, Africa would not equal Asia’s extent. Pace Merrills’s restriction of such an association of Africa and Europe to Sallust (Andrew Merrills, History and Geography in Late Antiquity (Cambridge: Cambridge University Press, 2005), 76–7). The broader currency of this model is demonstrated by its inclusion in Lucan’s excursus on the geography of Libya: ‘the third part of the world is Libya, … / but, if you go by the winds and sky / it will be a part of Europe’.36Lucan, 9.411–3, ed. Shackleton Bailey, 240, trans. Susan H. Braund (Oxford: Oxford University Press, 1992), 118 [hereafter Lucan]: ‘Tertia pars rerum Libye, … / at, si ventos caelumque sequaris, / pars erit Europae’. Though increasingly marginalised, the bipartite model was not entirely forgotten and a rhetorical opposition of Europe and Asia regained currency in the Augustan era.37Philip Hardie, Vergil’s Aeneid: Cosmos and Imperium (Oxford: Clarendon Press, 1986), 311–13. Nevertheless, looking towards the Middle Ages, it is the tripartite division which predominates.
This is substantially the picture we find among the Fathers in late antiquity. Augustine, Orosius and Isidore of Seville each present the same tripartition of Asia, Europe and Africa, while pointing towards others who adjoin Africa to Europe, creating an east-west bipartition of the world.38Augustine, De civitate Dei 16.17, ed. B. Dombart and A. Kalb, CCSL 47–8 (Turnhout: Brepols, 1955), 521; Augustine, Questionum in heptateuchum 6.15, ed. J. Fraipont, CCSL 33 (Turnhout: Brepols, 1958), 321; Orosius, 1.2.1, ed. Zangemeister, 5; Isi., Etym. 14.2.2. Similarly among pagan authors of this period as, e.g., Servius, In Vergili Aeneidos 1.385, ed. Thilo, 1:130. The solidification of this view in the fourth century should not surprise us, given the significance of Sallust and to a lesser extent Lucan as school texts around this time.39Patrick Gautier Dalché, ‘L’enseignement de la géographie dans l’antiquité tardive’, Klio 96, no. 1 (2014): 150–1; H. I. Marrou, A History of Education in Antiquity, trans. George Lamb (New York: Mentor, 1956), 373–5; Alan Cameron, Greek Mythology in the Roman World (Oxford: Oxford University Press, 2004), 3–4. Despite this ostensible agreement there is a small but significant change to the formula among all three Christian authors.40Wick, Lucanus, 160. For Greek and Roman geographers, the parts of the world were understood centrally in terms of the bodies of water that divided them.41There are exceptions, notably Varro, De lingua latina 5.5.31, ed. Goetz and Schöll, 11. See more broadly Nicolet, ‘Les “quartre sages”’, 180–1n63. Pomponius Mela provides a particularly clear example. After describing the ocean and its inlets, he notes simply that: ‘by this [our] sea and by two famous rivers, the Tanais and the Nile, the whole earth is divided into three parts’.42Pomponius Mela, De chorographia 1.8, ed. Silberman, 3, trans. F. E. Romer, Pomponius Mela’s Description of the World (Ann Arbor: University of Michigan Press, 1998), 36: ‘hoc mari et duobus inclutis amnibus, Tanai atque Nilo, in tres partes universa dividitur’. Cf. Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Sallust, Bellum Iugurthinum 17.3, ed. Kurfess, 68; Strabo, Geography 1.2.1, 2.5.18, trans. Jones, 1:49–51, 467–9. For the Fathers, however, rivers are no longer the key descriptors. Although they are still noted by both Isidore and Orosius, the three parts are described by these Christian authors first and foremost in terms of the cardinal points. Indeed, Augustine does not even mention river boundaries in the City of God, but simply explains that: ‘the part called Asia extends from the south, through the east, to the north; Europe extends from the north to the west; and then Africa begins and extends from the west to the south’.43Augustine, De civitate Dei 16.17, CCSL 48, 521, trans. R. W. Dyson (Cambridge: Cambridge University Press, 1998), 725: ‘ista, quae Asia nuncupatur, a meridie per orientem usque ad septentrionem peruenit; Europa uero a septentrione usque ad occidentem, atque inde Africa ab occidente usque ad meridiem’. While this may indicate, as Claudia Wick suggests, a more theoretical engagement with the world, treating it increasingly in abstraction, it also represents a turn towards a greater universalism in vision. As this new form of description was adopted centrally by the overtly Christian authors of late antiquity, it may also reflect the significance of the cardinal points in the Bible as descriptors of geographical universality.44E.g., Gn 13 and Lk 13:29. Compare the use of water boundaries in the not overtly Christian Martianus Capella, De nuptiis 6.622–3, ed. Willis, 217.
This particular insight warrants some further consideration as to the use of the cardinal points in geographical description through antiquity. Although they are certainly related both in antiquity and in the Middle Ages, their use is not obviously equivalent.45For a survey of medieval formulae, see Goetz, Gott und die Welt, vol. 1, bk. 2, 136–40; cf. Mauntel, Erdteile, 156, and Christoph Mauntel, ‘The T-O Diagram and its Religious Connotations: A Circumstantial Case’, in Geography and Religious Knowledge in the Middle Ages, ed. Christoph Mauntel (Berlin: De Gruyter, 2021), 79–80. For Roman authors, the four-part division was used as much for the heavens as for the earth. Pliny, for example, castigates the ancients for having only four winds for ‘as many parts of the world’, explaining that subsequent generations eventually settled on eight winds with two for each of the ‘four parts of the sky’.46Pliny, Naturalis historia 2.46.119, ed. Rackham, 1:260–1: ‘per totidem mundi partes’; ‘quattuor caeli partes’. Pomponius Mela also describes the four parts as general divisions of earth and sky, which are governed by the movement of the sun, and he links this to a discussion of the two hemispheres and five zones.47Pomponius Mela, De chorographia 1.3–4, ed. Silberman, 2; cf. Isi., Etym. 3.42. This seems to relate to the trend among Greek geographers, going back to Crates of Mallos and remaining at least marginally relevant throughout antiquity and the Middle Ages, according to which the globe contained four inhabitable landmasses, mirrored north-south and east-west.48Alfred Hiatt, Terra Incognita: Mapping the Antipodes Before 1600 (Chicago: University of Chicago Press, 2008), 16–17; Wesley M. Stevens, ‘The Figure of the Earth in Isidore’s “De natura rerum”’, Isis 71, no. 2 (1980): 269–70; Germaine Aujac, ‘L’image du globe terrestre dans la Grèce ancienne’, Revue d’histoire des sciences 27, no. 3 (1974): 205–10. See Martianus Capella, De nuptiis 6.604–7, ed. Willis, 211–12. As a result, the cardinal points typically described the entire world rather than just the oecumene. While they are occasionally used to denote regions of the oecumene, as for example in Agatharchides, this appears to have been an exception in the Roman world.49Agatharchides, De mari Erythraeo 64, trans. Karl Müller, Geographi Graeci minores, vol. 1 (Paris: Ambroise Firmin Didot, 1855), 156; see Nicolet, ‘Les “quartre sages”’, 180–1n63–5.
The use of the cardinal points for dividing the oecumene may have been pioneered by the Fathers, but there is another stage of late antique texts that crucially mediates between this patristic definition of the three parts and its medieval reception. One text in particular is of crucial significance here: the Cosmographia of Julius Honorius and Pseudo-Aethicus. The history of this text poses many problems. It dates to the fifth century and is attributed in the earliest manuscript to Julius Honorius, the only named author.50Nicolet, ‘Les “quatre sages”’, 162–3. On the complex manuscript tradition, the question of two ‘recensions’ and the relationship of Julius Honorius to Julius Orator, see Gautier Dalché, ‘Les “quatre sages”’, 184–92. It was circulating widely, at least in Italy, by the sixth century and is specifically recommended by Cassiodorus for his monks to read.51Gautier Dalché, ‘Les “quatre sages”’, 187–8; Cassiodorus, Institutiones 1.25, ed. R. Mynors (Oxford: Clarendon Press, 1961), 66. Although the earliest copy of the text is not obviously Christian, it gained popularity in Christian communities and copies of the text beyond the earliest manuscript show evidence of Christian additions.52These later copies lack any reference to Julius Honorius or Julius Orator and are attributed instead to Julius Caesar in earlier manuscripts: ‘Cosmographia Iulii Caesaris’ or ‘Cronica Iulii Caesaris’ (Alexander Riese, Geographi latini minores (Heilbronn: Henninger, 1878), 21). Besides Cassiodorus, the original title: ‘Cosmografia’ appears in Paris, Bibliothèque nationale de France, Latin 2769, 23v. On its relationship to Paris, Bibliothèque nationale de France, Latin 4808, see Gautier Dalché, ‘Les “quatre sages”’, 190, and E. A. Lowe, Codices Latini Antiquiores, 12 vols (Oxford: Clarendon Press, 1934–71), 5:10, no. 550. More importantly, unlike the previously discussed geographies, which all use the classical tripartite model to structure their texts, the Cosmographia instead describes the geography of the world according to the four ‘oceans’: north, south, east and west.
Despite its currency in Italy across the sixth and seventh centuries, the version of Julius Honorius lost popularity around the eighth century and was supplanted by a new version of the text. Later attributed to (Pseudo-)Aethicus, this new version consisted of a variant of Julius Honorius, to which an excerpt from Orosius’s geographical introduction had been added.53This is not to be confused with the Cosmographia of Aethicus Ister, edited by Michael W. Herren, The Cosmography of Aethicus Ister: Edition, Translation, and Commentary (Turnhout: Brepols, 2011), and Otto Prinz, Die Kosmographie des Aethicus, MGH QQ zur Geistesgesch. 14 (Munich: MGH, 1993). On the relationship of these texts, see Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 161n10 and 198–9. This is the version that would remain influential throughout the High Middle Ages. It was widely copied into geographical manuals, alongside a variety of other geographical texts, as for example in the collection, common to the north of France, that consisted of Jordanes’s Getica, Orosius’s History, Pseudo-Aethicus and the Antonine Itinerary.54Gautier Dalché, ‘Les “quatre sages”’, 192–8. In fact, it was in just such a compilation that Lambert of Saint-Omer read the Cosmographia.55Saint-Omer, Bibliothèque d’Agglomération, MS 717; Albert Derolez, The Autograph Manuscript of the Liber Floridus: A Key to the Encyclopedia of Lamber of Saint-Omer (Turnhout: Brepols, 1998), 193.
The addition of Orosian material and the manuscript context alongside Orosius puts a fine point on the schematic inconsistency that was left to the Middle Ages. For while a two-part division of the world could be easily squared with the three parts, it is not obvious how to harmonise the three- and four-part models.56A continued awareness of these conflicting models may be reflected in an inscription shared across three eleventh-century computus manuscripts: Dijon, Bibliothèque municipale, MS 448, 74v: ‘mundus quatuor angulos habet et partes tres diuisas’ (The world has four corners and three separate parts). Also Trier, Stadtbibliothek, Hs. 4° 1084/115, 99r and Boulogne-sur-Mer, Bibliothèque municipale, MS 188, 30r. For descriptions and reproductions, see Patrick Gautier Dalché, Mappae mundi (VIIIe–XIIe siècle), 2 vols (Turnhout: Brepols, 2024), Dij4, Boul2, Tri1 [hereafter Gautier Dalché, MM] (all citations follow Gautier Dalché’s alphanumeric code for each individual map). On these diagrams, see Obrist, ‘Wind Diagrams’, 58, and Bruce Eastwood, Ordering the Heavens: Roman Astronomy and Cosmology in the Carolingian Renaissance (Leiden: Brill, 2007), 416. Pseudo-Aethicus highlights this problem himself when moving from Julius Honorius to Orosius:
Those who measured near and far these quadripartite contents of the whole earth, our ancestors, established that it should be reckoned tripartite, investigating the entire orb surrounded by the border of the ocean sea, and they considered those three parts Asia, Europe and Africa. Although they were not lacking, who as we have said present two parts, Asia and Europe; they adjoin Africa, adding it into Europe, because it both lacks the latitudinal area and is woefully subjected to a region of the sky, labouring from its air, poisonous and tainted, full of innumerable beasts, frightful and unknown to mankind. But returning to the point, without prejudice to those who want all this together to be two parts, let us begin to relate the tripartite division.57Pseudo-Aethicus, Cosmographia 2.1, ed. Alexander Riese, Geographi latini minores (Heilbronn: Henninger, 1878), 90 [hereafter Ps.-Aeth.]: ‘Hanc quadripertitam totius terrae continentiam hi qui dimensi sunt longe citroque maiores nostri tripertitam reputari definierunt, investigantes universum orbem oceani maris limbo circumdatum, easque tres partes Asiam Europam et Africam reputarunt. Quamvis non defuerunt, qui duas partes sicut diximus perhiberent, Asiam et Europam; Africam vero in Europa adiciendam adfinierunt, quia et spatio latitudinis eget et caeli male subiacet climati, laborans aeribus suis, venenis †fucisque repleta inmanium et incognitarum humano generi innumerabilium bestiarum. Sed ad propositum remeantes absque illorum praeiudicio, qui hoc omne totum duas partes esse voluerunt, tripertitam divisionem dicere incipiamus’. This text is based closely on Orosius, 1.2.1, ed. Zangemeister, 5, trans. Fear, 36: ‘Maiores nostri orbem totius terrae, oceani limbo circumsaeptum, triquadrum statuere eiusque tres partes Asiam Europam et Africam uocauerunt, quamuis aliqui duas hoc est Asiam ac deinde Africam in Europam accipiendam putarint’ (Our ancestors divided the whole world, surrounded as it is by the belt of the Ocean, into three rectangular blocks, and called these three parts Asia, Europe, and Africa, although there are some who believe that there are two parts: namely Asia and Europe, including Africa in the latter part).
Pseudo-Aethicus not only acknowledges the issue here, but suggests a particular relationship between the three and fourfold divisions. The world seems to be properly divided into four and it is only considered (reputare) tripartite according to the geographical tradition of ‘our ancestors’. This privileges the four-part model as a natural arrangement of the world, an implication that may also be reflected in the trend of referring to three partes, but four climata.58Tres partes: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Lucan, 9.411, ed. Shackleton Bailey, 240; Isi., Etym. 14.2.1–2; Orosius, 1.2.1, ed. Zangemeister, 5; Ps.-Aeth., 2.1, ed. Riese, 90; HoSV, Desc. 7, ed. Gautier Dalché, 138; Hon., Im. mun. 1.7, ed. Flint, 52; Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 37; cf. Tres Climata: OV, 11.12, ed. and trans. Chibnall, 6:70–1; Quattuor climata: Isi., Etym. 13.1.3; Hon. Im. mun. 1.91, ed. Flint, 81; Bede, De natura rerum 10, CCSL 123A, 200–1; cf. Quattuor Partes in Pliny, Naturalis historia 2.46.119, ed. Rackham, 1:260–1; Quattuor Cardines in Orosius, 2.1.5, ed. Zangemeister, 35. See also Mauntel, Erdteile, 42–4. Pseudo-Aethicus’s answer to this dilemma is to simply distinguish the two different modes of geographical division and treat them separately, without attempting to combine or harmonise them.
In the Twelfth Century
Despite the popularity of the Cosmographia, the quadripartite division never took hold as a structural device for geographical texts. Due no doubt to their prominence among the authoritative ancient and patristic sources, Europe, Africa and Asia remained canonical throughout the Middle Ages. The four-part model of Julius Honorius nevertheless influenced the subsequent tradition and while Pseudo-Aethicus was happy simply to present these models alongside one another, the cordon sanitaire would not last. A major catalyst to this breakdown was the use of maps in education and their growing influence on geographical texts, especially into the twelfth century.59Patrick Gautier Dalché, ‘Maps in Words: The Descriptive Logic of Medieval Geography, from the Eighth to the Twelfth Century’, in The Hereford World Map: Medieval World Maps and their Context, ed. P. D. A. Harvey (London: British Library, 2006), 223–42, and Patrick Gautier Dalché, ‘Comment et pourquoi décrire une mappemonde au Moyen Âge?’, in Figures de l’autorité médiévale: Mélanges offerts à Michel Zimmermann, ed. Pierre Chastang, Patrick Henriet and Claire Soussen (Paris: Publications de la Sorbonne, 2016), 69–88. Indeed, the Cosmographia was itself already presented as excerpts from a sphaera and had been closely associated with the use of a map for geographical education by Cassiodorus.60Gautier Dalché, ‘L’enseignement de la géographie’, 158, 164; Cassiodorus, Institutiones 1.25, ed. Mynors, 66. From the early Middle Ages onwards the division between three- and four-part models slowly blurred as we find attempts to harmonise their content and corresponding modifications to the traditional tripartite model. So, although tripartition remained dominant, by the twelfth century the cardinal points were increasingly used autonomously to divide the world, either alongside or in conjunction with the three parts.
The broad influence of the Cosmographia can be seen in the reception of its narrative frame.61Nathalie Bouloux, ‘L’espace habité’, in Gautier Dalché, ed., La terre, 259–60. Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 157–218, esp. 203–5; Guido of Pisa, Liber Guidonis, ed. Michele Campopiano (Florence: SISMEL, 2008), 21–2; Lamb., Lib. flor., 241r (Gautier Dalché, MM, Gent13). Later manuscripts of Julius Honorius begin with a story about a survey of the entire world under Julius Caesar that was undertaken by four scholars, one for each of the four regions of the world: ‘when Julius Caesar and Mark Antony were consuls, the entire earth was scoured by four most wise and select men: Nicodemus the east, Didymus the west, Theodotus the north and Polyclitus the south’.62Jul. Hon., Cosm. 1, ed. Riese, 21: ‘Iulio Caesare et Marco Antonio consulibus omnis orbis peragratus est per sapientissimos et electos viros quattuor: Nicodemo orientis, Didymo occidentis, Theudoto Septemtrionalis, Polyclito meridiani’. On the inclusion of this story in later copies of Julius Honorius, see Gautier Dalché, ‘Les “quatre sages”’, 185–6. Pseudo-Aethicus further expands the role of Julius Caesar, describing his commission of the survey.63Ps.-Aeth., 1.1*–2*, ed. Riese, 72: ‘Itaque lulius Caesar bissextilis rationis inventor divinis humanisque rebus singulariter instructus cum consulatus sui fasces regeret, ex senatus consulto censuit omnem orbem iam Romani nominis admetiri per prudentissimos viros et omni philosophiae munere decoratos’ (And so Julius Caesar, inventor of the leap year and singularly learned in things human and divine, when he held his consulship, resolved on degree of the senate that the whole world now of Roman name be measured by men most skill and distinguished by every field of philosophy). This story is prominently included on a variety of mappae mundi. For example, a marginal inscription on the top right corner of the Ebstorf map notes that ‘Julius Caesar first prepared [a world map] after envoys were sent across the extent of the whole world’.64Hartmut Kugler, Die Ebstorfer Weltkarte, 2 vols (Berlin: Akademie, 2007), 1:42: ‘Quam [i.e. mappa mundi] Julius Cesar missis legatis per totius orbis amplitudinem primus instituit’. A far more prominent example is found on the Hereford Map. In the bottom left corner, there is a pictorial representation of the survey, in which the Cosmographia has been conflated with the biblical census: ‘Luke in his gospel [2:1]: there went out a decree from Caesar Augustus that all the world should be described’.65Scott D. Westrem, The Hereford Map: A Transcription and Translation of the Legends with Commentary (Turnhout: Brepols, 2001), 9, trans. Westrem: ‘Lucas in evvangelio: Exiit edictum ab Augusto Cesare, ut describeretur huniversus orbis’. As such, it is Augustus, depicted with a papal crown, who holds the fictive decree, complete with an imperial seal bearing his name, instructing the geographers to survey the world.66Westrem, Hereford Map, 9–11. But only three geographers receive Caesar’s instructions: Nicodemus, Theodotus and Polyclitus. This discrepancy is also noted in four blocks of red writing that border the map, where it is explained, drawing near-verbatim from Pseudo-Aethicus, that: ‘the world began to be measured by Julius Caesar: / The entire east was measured by Nicodemus. / The north and west were measured by Theodotus. / The southern part was measured by Policlitus’.67Westrem, Hereford Map, 3, my trans.: ‘A Julio Cesare orbis terrarum metiri cepit: A Nicodoxo omnis Oriens dimensus est. A Theodoco Setpemtrion et Occidens dimensus est. A Policlito Meridiana pars dimensus est’ (emphasis added). Cf. Ps.-Aeth., 1.1*–2*, ed. Riese, 72. Somewhere in the intervening centuries Didymus got lost and with him the careful division laid down by Pseudo-Aethicus.
This blending of three and fourfold divisions of the oecumene is already evident in the early twelfth century. A perfect example both of the problems of combining these models and of some of the possible solutions is found in Lambert of Saint-Omer’s apocalyptic compendia, the Liber Floridus (ca. 1120). Lambert’s geographical knowledge consists centrally of an idiosyncratic mixture of Orosius and Pseudo-Aethicus, no doubt a product of his having read both from the same manuscript: Saint-Omer, Bibliothèque d’Agglomération, MS 717.68Derolez, The Autograph Manuscript, 193. He strives to harmonise the two models, but does so differently on two different occasions. The most visually striking example is found near the beginning of the manuscript in Lambert’s T-O diagram of peoples.69Lamb., Lib. flor., 19r (Gautier Dalché, MM, Gent2); cf. Derolez, The Autograph Manuscript, 49. The peoples on this map are drawn directly from Pseudo-Aethicus, in substantially the same order, with the boundaries of the three parts as described by Orosius written in the margins. This created certain problems. First, Pseudo-Aethicus does not list peoples for the southern region of the world, saying that there are too many to count, so Lambert lists instead the provinces of Africa alongside the peoples of Asia and Europe.70Ps.-Aeth., 1.48, ed. Riese, 90. More significantly, to deal with the fourth region, Lambert has decided to divide Asia in two with another off-centre red line. There he begins the list with his own addition: ‘northern Asia has these peoples’, after which he provides the list from the northern ocean in Pseudo-Aethicus.71Lamb., Lib. flor., 19r: ‘Asia Septentrionalis habet gentes’; Ps.-Aeth., 1.38, ed. Riese, 87–8. The remainder of Asia begins simply ‘the east has’, and the west is left for Europe.72Lamb., Lib. flor., 19r: ‘Oriens habet’.
Near the other end of the compilation, Lambert attempts another harmonisation in his famous map of Europe, entitled: ‘Europa mundi pars quarta’.73Lamb., Lib. flor., 241r: ‘Europe, one fourth part of the world’. In the marginal text, he provides the story of Julius Caesar’s survey of the world, noting specifically that this region was measured by Theodotus and that while it is called ‘pars tercia’, it is in reality ‘quarta’.74Lamb., Lib. flor., 241r: ‘nominatur pars tercia, sed uere est quarta’ ([Europe] is called the third part, but it is really one fourth part). Note that Latin ordinals like quartus can mean both the fourth (in a series) and one fourth (of a whole). Even though the geography is of the west, Theodotus is the surveyor assigned to the north in Pseudo-Aethicus.75Ps.-Aeth., 1.2*, ed. Riese, 72. Lambert had himself already cut down the number of surveyors in his chronicle of the six ages, dropping the western surveyor Didymus, just like the Hereford map, and blocking north and west together.76Lamb., Lib. flor., 138v: ‘Deinde regna orientis et meridiani ac septentrionis occidentisque et prouintias insulasque per tres uiros prudentissimos uidelicet Nicodoxum et Pollyclitonem atque Theodotum metiri precepit’ (Then [Julius Caesar] ordered that the kingdoms of the east, south, and north and west as well as the provinces and islands be measured by three most learned men, namely Nicodemus, Polyclitus and Theodotus). Gautier Dalché, ‘Les “quatre sages”’, 203–5, 209. So while this map is rightly used as evidence for an increased geographical focus on Europe and an increased awareness of northern Europe in the High Middle Ages, there is also a tension here between Europe as northern and Europe as western, where it ambiguously inhabits both.77Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–70; cf. Fraesdorff, Norden, ch. 3.
Even for authors who are not explicitly attempting to harmonise the Cosmographia with a tripartite geography, the influence of its quadripartite system is still clearly felt. This can be seen both in the shifting boundaries of Asia, Europe and Africa, as well as in the developing use of the cardinal points to name regions of the world. Asia had long been understood as characteristically eastern and synonymous with the east.78E.g., Orosius, 1.2.2, ed. Zangemeister, 5; trans. Fear, 36: ‘Asia … per totam transuersi plagam orientis extenditur’ (Asia … extends across the entire east). Cf. HoSV, Desc. 7, ed. Gautier Dalché, 138, and Nicolet, ‘Les “quatre sages”’, 181. The other three directions, however, posed greater problems that will be illustrated through two case studies: first, on the boundaries of Africa and its increasing association with Egypt and Ethiopia in the south and second, on the ambiguous position of Britain and the Atlantic Archipelago between north and west in twelfth-century geography.
Africa, Egypt and Ethiopia
Although Africa, and particularly Ethiopia, had been considered characteristically southern since antiquity, the specific geography was never terribly clear. Already in Homer, the Ethiopians are imagined living at the edge of the world.79Romm, Edges of the Earth, 45–60. A direct equation between Africa and the south, however, never developed as it did for Asia, due no doubt to Africa’s own tenuously autonomous status and the lingering influence of ancient two-part models. What is more, the geographies of late antiquity are inconsistent about the exact location of Ethiopia in relation to Asia. Orosius tries to present a reasonably tidy picture that mirrors a T-O diagram of the world, but ultimately remains vague about the exact relationship of Asia and Africa at their extremities. He notes that the ocean lies to the south of both India and Egypt Superior. Egypt likewise lies to the east of southern Africa and Africa borders the ocean across the Ethiopian desert to the south. At the same time, in his rather confusing description of the Nile’s course, Orosius explains that the Nile emerges near the Red Sea and flows west then north. He notes, though, that some authors set its source near Mount Atlas, whence it flows underground into a great lake before turning east through the Ethiopian desert and then left into Egypt.80Orosius, 1.2.9, 15, 27–34, 87–92, ed. Zangemeister, 5–8, 13–14. But this now puts Ethiopia to the south of Egypt as well.
Isidore provides a somewhat different picture. As usual he situates Egypt in Asia and, like Orosius, sets Syria to its east, Libya to the west and the Mediterranean to the north. Its southern boundary is, however, uncharacteristically vague, extending to the interior all the way to Ethiopia. Africa likewise begins from Egypt, but runs south through Ethiopia to the west.81Isi., Etym. 14.3.27, 14.5.3. So Isidore appears to imagine an African Ethiopia wrapping around Egypt to the south. Finally, Julius Honorius and Pseudo-Aethicus situate African geography in the southern region of the world, beginning the list of southern provinces: ‘Egypt, Ethiopia, Africa’, but this south was of course never meant to conform to the world’s tripartition and contains other Asian provinces like Arabia as well.82Jul. Hon., Cosm. 43–4, ed. Riese, 47; Ps.-Aeth., 1.39–46, ed. Riese, 88: ‘Aegyptum, Aethiopiam, Africam’. So while Isidore’s placement of Ethiopia to the south of Egypt seems to have been more influential in the early Middle Ages, these descriptions caused evident confusion for medieval readers like Dicuil (ca. 825), who devotes an entire chapter of his geography to the relationship of Egypt and Ethiopia.83Dicuil, Liber de mensura orbis terrae prol.6, 4, ed. Tierney, 44–5, 54–7; cf. Situs orbis 4.13, ed. Patrick Gautier Dalché, ‘Situs orbis terre vel regionum: Un traité de géographie inédit du haut Moyen Âge (Paris, B. N. latin 4841)’, Revue d’histoire des textes 12–13 (1982–3): 167, and Ravennatis anonymi cosmographia 2.20–1, ed. Joseph Schnetz, Intineraria Romana, vol. 2 (Leipzig: Teubner, 1990), 32–3. Confusion around India’s relationship to Ethiopia also continued into the later Middle Ages. E.g., Gervase of Tilbury, Otia imperialia 2.3, ed. and trans. S. E. Banks and J. W. Binns (Oxford: Clarendon Press, 2002), 183; cf. Bernard Hamilton, ‘Continental Drift: Prester John’s Progress through the Indies’, in Prester John, the Mongols and the Ten Lost Tribes, ed. Charles F. Backingham and Bernard Hamilton (Aldershot: Variorum, 1996), 238–40, and Rouxpetel, L’Occident au miroir, 115–37.
By the twelfth century an alternative picture began to emerge as the south gained increasing autonomy as a distinct region of the world. We can see this already in the Imago Mundi (1110–39) of Honorius Augustodunensis.84Flint in Hon., Im. mun. 7–8; more generally on Honorius, see the collected essays in Valerie I. J. Flint, Ideas in the Medieval West: Texts and Their Contexts (London: Ashgate, 1988). For surveys on Honorius’s life see Karl Kinsella, ‘Edifice and Education: Thought in Twelfth-Century Europe’, 2 vols (PhD diss., University of Oxford, 2016), 1:214–18, and Fabian Schwarzbauer, Geschichtszeit: Über Zeitvorstellungen in den Universalchroniken Frutolfs von Michelsberg, Honorius’ Augustodunensis und Ottos von Freising (Berlin: Akademie, 2005), 40–2. His description of Africa follows his broader pattern of providing north and south boundaries only for intermediate sized regions, rather than to the individual provinces. So, for example, the central strip of Asia begins with India, which is given a northern and southern boundary, the Caucasus Mountains and the Red Sea, as well as an eastern and western boundary. Then for each subsequent region (Parthia, Mesopotamia and Syria), only an east and west boundary is provided, the Indus, Tigris, Euphrates and Mediterranean respectively.
This description of central Asia leads into one of Egypt. Honorius explains that ‘beginning in the east, these aforementioned regions extend in a straight line to the Mediterranean, and Egypt adjoins them to the south’.85Hon., Im. mun. 1.17, ed. Flint, 57: ‘Hee superius dicte regiones ab oriente incipientes, recta linea ad Mediterraneum mare extenduntur, quibus versus Austrum Egyptus connectitur’. Traditionally enough, Egypt forms the southern section of Asia, but only this northern boundary is provided. For east and west, Honorius follows Isidore, but removes Syria leaving only the Red Sea to the east and Libya to the west.86Cf. Isi., Etym. 14.3.27. This separation of Syria from Egypt not only highlights the southernness of the latter, but follows, at least indirectly, Pseudo-Aethicus’s placement of both.87Ps.-Aeth., 1.5, 43, ed. Riese, 73, 88. Flint cites no influence of the Cosmographia on the Imago Mundi (Flint in Hon., Im. mun. 46–7, passim). The use of the Red Sea as the sole eastern boundary may also reflect a cartographical representation of geographical space, given the prominence of the Red Sea in the south-east corner of most mappae mundi (fig. 2).88Gautier Dalché in HoSV, Desc. 171n43, and ‘Maps in Words’, 230–1. This is clearest on the Munich Map (Munich, Bayerische Staatsbibliothek, CLM 10058, 154v), but see also the Sawley Map (Cambridge, Corpus Christi College, MS 66, 1v). Gautier Dalché, MM, Mün27, Camb2. On these maps, see the chapters by Nathalie Bouloux, ‘The Munich Map’, and Alfred Hiatt, ‘The Sawley Map (c. 1190)’, in Terkla and Millea, eds, Critical Companion to English Mappae Mundi. Honorius also conspicuously extends the boundaries of Africa: ‘it arises on the east of the Indus river and inclining through the south it extends into the west’.89Hon., Im. mun. 1.30, ed. Flint, 63: ‘Hec in oriente Indi fluminis surgit, et per meridiem vergens, in occidentem tendit’. The earliest family of manuscripts, C, has ‘Nili’ in place of ‘Indi’. This would see Africa extend as far as India in the east, a description that while puzzling does bears a close resemblance to the relative location of the Indus and eastern tip of Ethiopia on contemporary mappae mundi.90This can be seen, e.g., on both the Munich and Sawley Maps (Munich, Bayerische Staatsbibliothek, CLM 10058, 154v, and Cambridge, Corpus Christi College, MS 66, 1v). Even if this is an error and we follow the earliest family of manuscripts that gives the Nile instead, Honorius’s phrasing still notably mirrors Isidore’s description of Asia extending ‘from the south through the east up to the north’.91Isi., Etym. 14.2.2, my trans.: ‘Asia a meridie per orientem usque ad septentrionem pervenit’. The implication here should be clear as Isidore specifically notes in the next sentence that Asia takes up half the world.92Isi., Etym. 14.2.3. Following Isidore once again, Honorius describes Ethiopia as ‘toward the south’ and divided into two parts, ‘one in the east … the other in the west’, beyond which lie great deserts and the ocean.93Hon., Im. mun. 1.32, ed. Flint, 64: ‘Versus meridiem vero est Ethiopia, ab Etham dicta, una in oriente … altera in occidente … Ultra Ethiopiam sunt maxima loca deserta … Deinde est maximus oceanus’. Thus, although he has deftly fit everything into a largely traditional mould, Honorius has conspicuously extended the boundaries of Africa across the southern region of the world and associated it more clearly with the south.
While we can still describe Honorius’s solution to the southern dilemma as traditional, the same cannot be said for the Descriptio mappe mundi (1130–5) of Hugh of Saint Victor.94Gautier Dalché in HoSV, Desc. 55–8. On Hugh more generally, see Paul Rorem, Hugh of Saint Victor (Oxford: Oxford University Press, 2009), ch. 1. Like the Fathers, Honorius primarily used the cardinal points to describe the three parts, but Hugh goes a step further. At the end of his discussion of Africa, he has a brief aside where he explains that there are a number of different names given to this region: ‘it should be known that writers sometimes set down Libya for Africa and Africa for Libya, but Ethiopia, Egypt, Libya and Africa are also set down indifferently for the southern region’.95HoSV, Desc. 18, ed. Gautier Dalché, 151: ‘sed sciendum quod scriptores aliquando Libiam pro Affrica, et Affricam pro Libia ponunt, sed et Ethiopiam, Egyptum, Libiam, Affricam pro meridiana plaga indifferenter ponunt’. What was perhaps implicit in Honorius is now unambiguously stated. While this may appear an anodyne terminological clarification, it has far-reaching implications, as we see in Hugh’s most radical departure from the ancient and medieval geographical tradition: his inclusion of Egypt within Africa. Radical as it may be, Hugh is not so out of step with his age as he may first appear.96Cf. Gautier Dalché in HoSV, Desc. 51n34, 171n43. The equation of Africa with south can be found already in the B family of glosses on Martianus Capella associated with the school of Remigius of Auxerre (d. ca. 908).97Paris, Bibliothèque nationale de France, Latin 8674, 71v, and Paris, Bibliothèque nationale de France, NAL 340, 77v: ‘Lybicum i.e. Africum i.e. australem’. On Martianus Capella glosses, see Lozovsky, Earth Is Our Book, 113–38, esp. 116, and Cora E. Lutz, ‘Martianus Capella’, in Catologus translationum et commentorum, vol. 2, ed. Paul Oskar Kristeller (Washington, DC: Catholic University of America Press, 1971), 372–6. Neither is he the only author to draw such a connection between Africa and Egypt. Both Peter the Venerable and Guibert of Nogent similarly list Egypt alongside Africa, Libya and for Guibert Ethiopia too, as distinguished from ‘the east’.98Peter the Venerable, Liber contra sectem Saracenorum prol., D 179vs, ed. James Kristzeck, Peter the Venerable and Islam (Princeton: Princeton University Press, 1964), 226 (emphasis added): ‘toto fere ut dictum est armis oriente subacto, Egyptum, Lybiam, Affricamque uniuersam prophanae religioni subiecit, et sic duabus mundi partibus occupatis’ (see p. 000 below for a translation); Guibert, Dei gesta 1.4, CCCM 127A, 98–9: ‘Huius nefariae institutionis obscuritas christianum tunc nomen obtexit et adhuc per Orientis pene universi, Affricae, Egypti, Ethiopiae, Libiae et iuxta nos Hispaniae remotissimos sinus oblitterat’ (The shadow of this wicked sect covered the Christian name and has now consigned it to oblivion through the furthest recesses of almost the whole east, of Africa, Egypt, Ethiopia, Libya and of Spain, next to us). This has further ramifications for the geography of Asia. Africa no longer serves as the southern boundary of eastern and western Asia and there is no longer a southern Asia.99HoSV, Desc. 9–12, cf. capitula, ed. Gautier Dalché, 140–5, cf. 133. N.B. ‘IX. De regionibus et gentibus Asye ad austrum’ is listed among the capitula, however in the text we find ‘IX. De prouinciis et ciuitatibus Asie ad ortum solis’, covering the geography of India, which includes Ethiopian India ending at the Red Sea in the south (on the regions and peoples of Asia to the south; on the provinces and cities of Asia to the east). The Red Sea therefore has become the boundary of Asia and Africa, with both Ethiopia and Egypt described as lying to its west.100HoSV, Desc. 9–10, 14, ed. Gautier Dalché, 140–1, 146.
The Descriptio is not Hugh’s only geographical work, even if it is his most extensive. He also included a geographical section in his Chronica (ca. 1130), and while it is somewhat more traditional than the Descriptio, there are still important similarities.101On the Chronica, see Mortensen, ‘Hugh of St. Victor on Secular History’, 3–30; Green, ‘De Tribus Maximis’, 484–93. For a complete transcription of the Chronica, see Hugonis de Sancto Victore operum Editio auspiciis Gilduini abbatis procurata et IV voluminibus digessa, ed. Rainer Berndt and José Luis Narvaja, vol. 1 (Münster: Aschendorff, 2017). As Berndt and Narvaja reproduce a single manuscript: Paris, Bibliothèque nationale de France, Latin 15009, all my citations follow that manuscript. Here the entire south-eastern section of the world, from eastern Ethiopia to Upper Libya(!), is clearly distinguished from the provinces of Africa. But its relationship to Asia is also at least somewhat ambiguous:
On the southern side of Asia, the Red Sea extends from the eastern ocean towards the west up to the boundaries of Arabia and Palestine. / Between the Red Sea and the southern ocean to the east, above, is Upper Ethiopia, which is bounded on the west by the Ethiopian mountains along its side. / Underneath eastern Ethiopia these regions are extended to the west: Upper Egypt, the Egyptian desert, Lower Egypt, Upper Libya / These regions are in Africa: Ethiopian Nadaberes …102Paris, Bibliothèque nationale de France, Latin 15009, 16r: ‘In latere australi asie porrigitur mare rubrum ab occeano orientali occidentem uersus usque ad confinia arabie et palestine. / Inter mare rubrum et occeanum australem ad orientem supra ethiopia est superior que ab occidente montibus ethiopicis per transuerssum terminatur. / Sub ethiopia orientali ad occidentem hee sunt regiones porrecte / Egiptus superior / Deserta egypta / Egyptus inferior / Lybia superior / In affrica sunt regiones iste / Ethiopes nadaberes [etc.]’.
Unlike the north of Asia or the north of Europe, it is not the provinces of southern Asia that stretch out (porrigere), but the Red Sea. As a result, Upper Ethiopia is not listed as a region here, or distinguished as such in the manuscript like Scythia or Provence are in similar contexts.103Cf. Leipzig, Universitätsbibliothek, MS 350, 107v, and Munich, Bayerische Staatsbibliothek, Clm 2, 101v. The prominence of the Red Sea is further underscored by a complete lack of reference to the Nile as a geographical boundary, in contrast with the Tanais and Mediterranean. Finally, the division between Africa and Asia is itself deemphasised, occurring almost as part of the list of provinces under eastern Ethiopia and not associated with any geographical boundary. Therefore, even as Hugh conforms to the traditional tripartition here, the south once again stands out as an unusually coherent geographical space, bounded by the Red Sea in the east, the Atlantic Ocean in the west and centrally associated with Africa, Egypt and Ethiopia.
The increased prominence of the Red Sea as a geographical boundary can be found more broadly in the twelfth century. In the Liber Guidonis (1118–19), Guido of Pisa likewise rejects Egypt as the eastern boundary of Africa in favour of Arabia and the Red Sea.104Guido of Pisa, Geographica 122, ed. Schnetz, 140: ‘Africa habet fines ab oriente praelatam Arabiam et regionem maris Rubri et ex parte regionem Hebraeorum’ (Africa has as boundaries on the east the aforementioned Arabia and the region of the Red Sea as well as in part the region of the Hebrews). Cf. Orosius, 1.2.8, ed. Zangemeister, 5; Isi., Etym. 14.5.3. On Guido, see Campopiano, ‘Construction of the Text’, 67–73; Campopiano in Guido of Pisa, Liber Guidonis, LI–LXII. So although Guido situates Egypt within Asia, Africa has now adopted Egypt’s traditional eastern boundaries, leaving Egypt’s exact position in south-western Asia rather unclear.105Guido of Pisa, Geographica 121, ed. Schnetz, 139; cf. Orosius, 1.2.27–34, ed. Zangemeister, 7–8; Isi., Etym. 14.3.27. We can also read Honorius’s removal of Syria from the eastern boundary of Egypt in the same light, since the position of Egypt as the explicit boundary of Asia and Africa has all but dropped out of the picture. This fundamentally underscores Gautier Dalché’s suggestion that Hugh’s inclusion of Egypt in Africa is a product of the prominence of the Red Sea on medieval maps, but as I have argued here, this is part and parcel of a much wider group of developments, central to which is the increasing autonomy of the south as a distinct geographical region.106Gautier Dalché in HoSV, Desc. 171n43.
Cartographic Representations of Africa
The expansion of Africa and the newly prominent position of the Red Sea are further underscored in cartographical developments down to the twelfth century. Although the early history of medieval maps is hazy, there is good reason to believe that the original ‘“mise en cart” of geographical texts’, which established many of the standard features of medieval maps, occurred during late antiquity, prior to our earliest extant maps of the eighth and ninth centuries.107Patrick Gautier Dalché, ‘L’héritage antique de la cartographie médiévale: Les problèmes et les acquis’, in Cartography in Antiquity and the Middle Ages: Fresh Perspectives, New Methods, ed. Richard Talbert (Leiden: Brill, 2008), 36–53, at 51: ‘“mise en cart” des textes geographiques’; Patrick Gautier Dalché, ‘Eucher de Lyon, Iona, Bobbio: Le destin d’une mappa mundi de l’antiquité tardive’, Viator 41 (2010): 1–22. From around the turn of the twelfth century, emanating particularly from northern France, many of the formal aspects of the great mappae mundi – such as the circular form, eastern orientation and quantity of miscellaneous details – began to solidify.108See p. 000 above. Within this process an analogous set of recurrent tensions appear around the relationship of Ethiopia and Asia, almost certainly developing in relationship with the textual tradition discussed above. There are two comparisons that will serve to highlight the cartographical tension: the representation of the Nile up to the end of the twelfth century and the continued tension around the border of Africa and Asia in the Ebstorf and Hereford maps.
Although a broadly standard depiction of the two Nile rivers had emerged by the thirteenth century, there was considerable variation up to at least the mid-twelfth century. In many pre-eleventh-century maps, the depiction of a single Nile is closely related to a strict delimitation of Africa to the south-west section of the world, in line with the T-O diagram and the sections of Orosius that are sympathetic to such a picture. The earliest extant mappa mundi, the Albi world map, presents just this sort of picture.109Albi, Bibliothèque municipale, MS 29, 57v; ed. F. Glorie, Itineraria et alia geographia, CCSL 175 (Turnhout: Brepols, 1965), 467–71. Gautier Dalché, MM, Alb1. On the Albi map, see Emmanuelle Vagnon and Sandrine Victor, eds, La Mappa mundi d’Albi: Culture géographique et représentation du monde au haut Moyen Âge (Paris: Éditions de la Sorbonne, 2022). It depicts the world in the shape of a sling wrapped around the Mediterranean and, while Asia, Europe and Africa are not labelled, we may reasonably infer their divisions.110On the form, see Anca Dan, ‘La mappemonde d’Albi – Un pinax chôrographikos’, Revue du Comité français de cartographie 234 (2017): 27; Gautier Dalché, ‘L’enseignement de la géographie’, 163; and Priscian, Periegesis 10–13 (ed. Paul van de Woestijne, La Périégèse de Priscien: Édition critique (Bruges: De Tempel, 1953), 37). The Caspian Sea and Red Sea are almost directly aligned and the Red Sea is depicted as (or right beside) the head of the Nile. The position of Ethiopia is therefore clear in this presentation. It covers the southern half of Africa, terminating alongside Africa at the Red Sea. A minor ambiguity arises around the position of Egypt. Orosius describes Africa as beginning from the border of Egypt and Alexandria, but sets Egypt to the west of the Red Sea.111Orosius, 1.2.8, 34, ed. Zangemeister, 5, 8. On the influence of Orosius and Pseudo-Aethicus, see Konrad Miller, Mappaemundi: Die ältesten Weltkarten, vol. 3, Die kleineren Weltkarten (Stuttgart: Jos. Roth, 1895), 58–9, but see also the more nuanced discussion in Jean-Baptiste Amat, ‘La Mappa mundi d’Albi au regard de la Mappa mundi du Vatican comme illustration de la nécessaire comparaison des mappae mundi du haut Moyen Âge’, in Vagnon and Victor, eds, La Mappa mundi d’Albi, 180–1, 189–94. The map has therefore labelled Egypt and Alexandria on opposite sides of the Nile, so that Egypt lies on the correct side of the Red Sea. As a result, unlike in Orosius’s description, Egypt is depicted as bordering on Ethiopia to the south. Nevertheless, the map carefully delimits Africa to the south-western section of the world, in line with Anca Dan’s suggestion that it follows the model of an Orosian orbis triquadrus.112Dan, ‘La mappemonde d’Albi’, 23–7.
The early Beatus maps are even clearer on this point. In three of four tenth-century Beatus manuscripts we find the same presentation of a single river Nile along with the clear demarcation of both Africa and Ethiopia to the west of Asia.113For the stemma, see John Williams, ‘Isidore, Orosius and the Beatus Map’, Imago Mundi 49, no. 1 (1997): 10. Cf. Valladolid, Biblioteca Universitaria, MS 433, 36v–37r; Seu d’Urgell, Museo Diocesano, Num. Inv. 501, VIv–VIIr. The schematic form is clearest in the earliest example, the Escalada Beatus.114New York, Morgan Library, MS M.644, 33v–34r. Reproduced in Williams, ‘Isidore’, 9. This map labels the three parts of the oecumene, which are schematically divided by the (unlabelled) Mediterranean in the centre, the Tanais, which runs directly north except to divide in two around the Riphaean Mountains, and the Nile, which curves sharply west. The term Asia runs across both pages, with two letters on either, leaving no ambiguity about the location of paradise, on the top left of the right page, nor about its boundary with Africa. Ethiopia, however, seems to disrupt the standard tripartition of the world, as its label is the same size as that of Europe, Asia and Libya, and it is clearly divided from Libya by the Nile. It is also cut off from Asia by the Taurus Mountains and Mount Lebanon. But, once again, the position of Egypt is ambiguous: not only is it nestled alongside Africa between the Nile and Mediterranean, but the last letter of ‘Libia’ crosses the mountain range which divides Egypt and Alexandria from the province ‘Africa’. The Saint-Sever Beatus (fig. 1) reveals a similar set of tensions.115Paris, Bibliothèque nationale de France, Latin 8878, 45bisv–45ter. Reproduced in Williams, ‘Isidore’, 12. A prominent river Nile runs directly south dividing Africa from Asia and Ethiopia. Both ‘Africa’ and ‘Libia’ are labelled overlapping across western Africa, but Lower Egypt and Alexandria, both situated on the western bank of the Nile, lie between the letters L and I of ‘Libia’ and are only divided from ‘Africa’ by a small river. While the Saint-Sever Beatus does at least label Upper Egypt on the Asian side of the Nile, these maps clearly depict Egypt either in whole or in part as part of Libya.
Content - style instruction: <INSERT FIGURES 1 AND 2 NEAR HERE>
We can contrast this presentation of a clearly western Africa, albeit one that still contains some ambiguities around the location of Egypt, with another group of maps on which the extent of Africa becomes even less clear. Beginning with the eighth-century Vatican map (Vat. lat. 6018), we can see an early example of the portrayal of the second Nile running across the southern edge of the world, a trend that would come to dominate high medieval maps.116Vatican, Biblioteca Apostolica Vaticana, Vat. lat. 6018, 63v–64r; ed. Glorie, CCSL 175, 454–66; Gautier Dalché, MM, VatV17. Reproduced in Williams, ‘Isidore’, 16. On the Vatican Map, see Edson, Mapping Time and Space, 61–4, and Amat, ‘La Mappa mundi d’Albi’. Unlike the previous set of maps, the Vatican map presents three visually distinct landmasses, separated primarily by an expansive Mediterranean in the centre, which links up with the Red Sea to divide the southern and eastern sections. These do not line up with the traditional tripartition of the world, however, since the south contains not only Africa and Ethiopia, but also Egypt.117CCSL 175, 458–9: ‘Ereptus Superior’, ‘mons Climax’ and ‘diserta Egypti’. While Africa and Asia are not labelled, they could be divided by ‘Mount Catabathmon’ (mons Cathal), which runs north-south from the Mediterranean to the second Nile (cf. fig. 2).118CCSL 175, 462; cf. Orosius, 1.2.9 (ed. Zangemeister, 5). There is no such demarcation to the south of the second Nile, which runs westward separating Ethiopia in the south from Africa and Egypt to the north. So, while the south-east corner beyond the second Nile is labelled ‘Egyptian Desert’, there is nothing to distinguish it from the Ethiopian desert next to it and only a standard provincial boundary-line distinguish these from ‘Libia-Grammantium’ to the west.119CCSL 175, 462–3. The toponyms, therefore, create ambiguity as to the extent of Ethiopia under the south of Asia and its relationship with Africa and Egypt. Indeed, given the east-west orientation of the Red Sea and the Nile no longer serving as a boundary of Africa, we are left with a strikingly Hugonian picture of a ‘south’ constituted by Egypt, Africa and Ethiopia, and demarcated geographically by the Red Sea.
The ambiguity created by a second Nile is not reserved to the Vatican map, but is found across a number of eleventh- and twelfth-century maps, as on the Cotton (fig. 3) and Munich maps (fig. 2) or the Tournai map of Palestine. The Cotton map (ca. 1025–50) further divides the second Nile into two sections, one in the east and one in the far west, following Orosius’s confusing discussion of the Nile flowing underground and emerging near the ‘fons Dara’, which is labelled in the bottom right.120London, British Library, Cotton MS Tiberius B V, 56v; reproduced in Miller, kleineren, 33, and transcribed in Gautier Dalché, MM, Lon17, at 1:194: ‘Dara source’. On the Cotton Map, see Appleton, ‘Northern World’, and Alfred Hiatt, ‘An Anglo-Saxon world view? The Cotton mappa mundi Reassessed’, Viator 51, no. 2 (2020): 93–129. Orosius, 1.2.31 (ed. Zangemeister, 7). I interpret the second river as an extension of the Nile, due to its visual similarity and the interpretation of Orosius. On the Munich map, a separate Dara river (‘fluuius Dara’) is labelled flowing from Mount Atlas. At first glance, the map appears to clearly divide Africa from Asia, with lower Egypt and Alexandria both situated to the north-east of the first Nile and set alongside Lake Chalearzum, which Orosius lists alongside Catabathmon on the border of Africa.121Orosius, 1.2.9, ed. Zangemeister, 5; Gautier Dalché, MM, 1:194: ‘lacus Calearsum’. Egypt likewise extends south-east to mount Climax, which straddles the second Nile. Once we look beyond the second Nile, however, this division becomes less straightforward. The map again seems to follow Orosius, who explains that after Lake Chalearzum, Africa extends through the Ethiopian desert to the southern ocean. Following the now eastern orientation of the Nile, we find just such a set of Ethiopian toponyms extending east of Mount Climax along the southern ocean. First are the barbarian Getuli, who Orosius sets in Africa to the south of the province of Tripolitana, near the Ethiopian ocean.122Orosius, 1.2.90, ed. Zangemeister, 13; cf. Isi., Etym. 9.2.15, 9.2.118–19; Gautier Dalché, MM, 1:194: ‘hic barbari Getuli’. Next we find Ethiopian Libya (‘Libia Ethiopum’), a region that is labelled twice, once here to the east and again to the south-west of Lake Chalearzum between the two parts of the second Nile. Finally, the Ethiopian desert lies south of Upper Egypt. Given both the presence of a canonically African nation and the express linkage of east and west through Ethiopian Libya, we can see how Africa’s southern extent has been wrapped eastward under Asia.
Content - style instruction: <INSERT FIGURE 3 NEAR HERE>
The Tournai map of Palestine presents a different approach and finds a clever way to mediate this issue. It comes from the late twelfth century, but is almost certainly based on an insular world map of the sixth to eighth century.123London, British Library, Additional MS 10049, 64v; reproduced in P. D. A. Harvey, Medieval Maps of the Holy Land (London: British Library, 2012), 48–9. On the Tournai maps, previously called the ‘Jerome’ maps, see Harvey, Holy Land, 40–59, and Gautier Dalché, ‘Eucher de Lyon’, 6–9. The two Niles are depicted on the right side of the map, but only their Asian section, as this is of course a regional map of Palestine. The second Nile therefore extends beyond the bottom of the map at the border of Africa, represented by the Catabathmon mountains, which only run between the two Niles like on the Vatican and Munich maps (fig. 2).124Munich, Bayerische Staatsbibliothek, Clm 10058, 154v; Vatican, Biblioteca Apostolica Vaticana, Vat. lat. 6018, 63v–64r. Upper Egypt is labelled three times across both sides of the first Nile and to the east of the source of the Nile is ‘Egyptian India’.125London, British Library, Additional MS 10049, 64v: ‘India Egipti’. The second Nile is here called the Nuchul, presumably following Pomponius Mela’s description of Ethiopia, which of course extends along its southern bank.126Pomponius Mela, De chorographia 3.96, ed. Silberman, 93; cf. Orosius, 1.2.31, ed. Zangemeister, 7, and Munich, Bayerische Staatsbibliothek, Clm 10058, 154v. On this map, however, Ethiopia has been specifically associated with Asia. Both in the west, across from Catabathmon, and in the middle we find ‘Egyptian Ethiopians’ beyond the second Nile. This serves to frame the depicted region of Ethiopia as part of Egypt, or at least assigned to Egypt, thereby drawing this otherwise ambiguous section of Ethiopia explicitly into Asia.127London, British Library, Additional MS 10049, 64v: ‘Ethiopes Egiptii’. Three further labels in the east reinforce this point: ‘Egyptian India’, ‘Ethiopians’ and ‘Ethiopian India’.128London, British Library, Additional MS 10049, 64v: ‘India Egipti’, ‘Ethiopes’ and ‘India Ethiopie’. Given the late antique origin of both the tripartition of India and the early model for this map, the presence of these features on an earlier exemplar cannot be ruled out.129Rouxpetel, L’Occident au miroir, 116–17; Gautier Dalché, ‘Eucher de Lyon’, 13. On the tripartition of India, see generally Marianne O’Doherty, The Indies and the Medieval West: Thought, Report, Imagination (Turnhout: Brepols, 2013), 33–8. The tripartition of India, however, gains popularity from precisely this period, in concert with increasing awareness of a plurality of eastern Christians among the Indians, Ethiopians and Nubians.130Rouxpetel, L’Occident au miroir, 104–5, 117n64; von den Brincken, Nationes Christianorum orientalium, 246, 264–6, 338–9, 420–7. Likewise, Paul Harvey identifies a possible shift from ‘India’ to ‘Egyptian India’ between the second and third recension of the Palestine map.131Harvey, Holy Land, 47–9. As such, rather than denoting specific geographical locations, the terms India and Ethiopia may serve to describe merely liminal eastern and southern locations as a means to clarify the problem of south-eastern geography.
Content - style instruction: <INSERT FIGURE 4 NEAR HERE>
As we can see, the problematic relationship of Egypt and Ethiopia and the corresponding issue around the border of Africa and Asia is not isolated to texts, but remained entangled with and reinforced by the continued problem of depicting this relationship cartographically. This tension remains unresolved across the High Middle Ages and moving into the thirteenth century we find a beautiful microcosm of the continued uncertainty around this boundary on the Hereford and Ebstorf maps. Unlike the previously discussed maps, the Hereford map (fig. 4) specifically labels the boundary of Africa and Asia in just the same way as the Sawley map.132Cambridge, Corpus Christi College, MS 66, 2. The Ethiopian Mountains emerge from the bend in the first Nile where it flows into the Mediterranean, and these mountains continue south for a short distance before turning east and converging once again on the Nile. Near the bend in this mountain range ‘the boundary of Asia and Africa’ is labelled in red, beyond which lies the ‘land of Ethiopia’.133Westrem, Hereford Map, 183, 185, my trans.: ‘Terminus Asye et Affrice’; ‘Terra Ethiopie’. Cf. Westrem’s discussion of the ambiguity of this region, which is also highlighted in the scribe’s confusion over Pliny’s measurement of Africa (Westrem, Hereford Map, 182, 340). This creates a unified southern region, extending under Asia up to and potentially including Nubia below the Red Sea, thereby extending the boundary of Africa along the south-eastern edge of the world. The Ebstorf map, by contrast, expressly restricts Africa to the south-west. Much like on the Tournai map, the Catabathmon mountains have been depicted running directly south from the mouth of the first Nile and in this case, the map goes a step further by specifying that ‘Africa begins in these mountains’.134Kugler, Ebstorfer Weltkarte, 1:96: ‘Mons Gathabathmon. In his montibus Affrica sumit initium’. Similar again to the Tournai map, potential references to Ethiopia in the south-east have been replaced with Egypt. Thus, south of Meroe we find the ‘Egyptian desert’ and its Satyrs, rather than an Ethiopia desert.135Kugler, Ebstorfer Weltkarte, 1:68: ‘Egyptia deserta’. Finally, at the eastern end of this region, there is a note drawn from Solinus about monstrous races that live between Egypt, Ethiopia and Libya.136Kugler, Ebstorfer Weltkarte, 1:68: ‘In hoc latifundio et inter Egyptum et Ethyopiam et Libiam sunt genera simiarum silvestria …’ (In this region between Egypt, Ethiopia and Libya there are wild species of monkey …). Cf. C. Julius Solinus, Collectanea rerum memorabilium 27.55–60, ed. Theodor Mommsen (Berlin: Weidmann, 1895), 127–8. This statement has, however, been recontextualised by another label right underneath it: ‘peoples of Egypt’.137Kugler, Ebstorfer Weltkarte, 1:68: ‘Gentes Egyptie’. Thus the ambiguity around this region remains through the thirteenth century, as do continued attempts to resolve the relationship of Africa and Asia, Ethiopia and Egypt.
Britain and the Islands of the North-Western Ocean
Moving to the other end of the world, as the south gained additional autonomy and began to be linked with one of three parts of the world, the position of Europe between north and west grew more ambiguous. This has been evaluated most often in terms of the expansion of northern European geography between the eighth and thirteenth centuries.138See Patrick Gautier Dalché, ‘La connaissance des régions septentrionales de l’Europe à l’époque carolingienne: un changement de point de vue’, Geographia Antiqua 25 (2016): 3–15; Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–71; Fraesdorff, Norden, 291–5; Wolever, ‘Post-Carolingian Geographies’, 153–63; and Thomas Foerster, ‘Imagining the Baltic: Mental Mapping the Works of Adam of Bremen and Saxo Grammaticus, Eleventh-Thirteenth Centuries’, in Imagining Communities on the Baltic Rim, from the Eleventh to Fifteenth Centuries, ed. Wojtek Jezierski and Lars Hermanson (Amsterdam: Amsterdam University Press, 2016), 37–58. There is a dual process at work here. European geography was not only expanding to include modern Scandinavia, central and eastern Europe, and the Baltic, but we also see a reorganisation of Roman geography as the received Roman north was progressively elided with the new medieval north. Central to the late Roman conception of the north was the Danube, which served as the fundamental dividing line between the Roman world and the barbaricum.139Orosius, 1.2.54, ed. Zangemeister, 9; Isi., Etym. 14.4.3. Yet in the Middle Ages, as new regions of the north entered the Latin awareness, they were often identified with Roman provinces along the Danube like Dacia, Suavia and Raetia.140Alheydis Plassmann, Origo gentis: Identitäts- und Legitimitätsstiftung in früh- und hochmittelalterlichen Herkunftserzählungen (Berlin: Akademie, 2006), 248–9; Leonid S. Chekin, Northern Eurasia in Medieval Cartography: Inventory, Text, Translation, and Commentary (Turnhout: Brepols, 2006), 22–5, 221, 248; Evelyn Edson, ‘Dacia ubi et Gothia: Die nordöstliche Grenze Europas in der mittelalterlichen Kartographie’, in Baumgärtner and Kugler, eds, Europa im Weltbild, 173–89; Wolever, ‘Post-Carolingian Geographies’, 153–63. See also the analogous issue around the relationship of Thule and Iceland: Dale Kedwards, The Mappae Mundi of Medieval Iceland (Cambridge: D. S. Brewer, 2020), 127–40. Thus, for example, it is entirely typical that in his account of the Danish invasions of England, written around 1150, Alfred of Beverley records a pagan army sailing from the Danube, which he presents as a province set alongside Norway, Sweden, Gothia and Germania.141Alfred of Beverley, Historia 7, cf. 6, ed. J. P. T. Slevin and trans. L. Lockyer (Woodbridge: Boydell, 2023), 115, cf. 103: ‘iterum pagani de Danubia, Nowegia, Suethia, Gethia, Germania et aliis nationibus cum .viii. regibus et .xx. comitibus nauigio Britanniam aduecti’ (Once again pagans from the Danube, Norway, Sweden, Gothia, Germany and other nations arrived by ship in Britain, with eight kings and twenty earls). I have lightly edited Lockyer’s translation. The ultimate source for ‘Danubia’ is Asser’s De rebus gestis Aelfredi 21, ed. William Henry Stevenson, Asser’s Life of King Alfred (Oxford: Clarendon Press, 1959), 19, trans. Simon Keynes and Michael Lapidge, Alfred the Great: Asser’s Life of King Alfred and Other Contemporary Sources (London: Penguin, 1983), 74: ‘Et eodem anno magna paganorum classis de Danubia Britanniam advenit’ (In the same year, a great Viking fleet arrived in Britain from the Danube …). It is impossible to say whether this term is original to the ninth century or a corruption of the turn of the eleventh century introduced into the only (no longer) extant manuscript, but it cannot postdate this as it is already found in Byrhtferth of Ramsey, Historia Regum 4.6, ed. Michael Lapidge (Oxford: Clarendon Press, 2023), 132–3. As a result, this terms appears to be restricted to English chroniclers writing about the great pagan army, e.g. John of Worcester, Chronicon 888, ed. Reginald R. Darlington and Patrick McGurk, trans. Jennifer Bray and Patrick McGurk, The Chronicle of John of Worcester, vol. 2, The Annals from 450–1066 (Oxford: Clarendon Press, 1995), 280; Liber Eliensis 1.39, ed. E. O. Blake (London: Royal Historical Society, 1962), 53; and Roger of Howden, Chronica 866, ed. William Stubbs, 4 vols (London: Longmans, 1868–71), 1:38. This phenomenon was already identified by the Bollandists in their edition of the life of St. Æthelthryth, drawn from the Liber Eliensis, and subsequent scholars have generally shared their conclusion that this must be a corruption of a term for Danes or Denmark (Jean-Baptiste du Sollier, Jean Pien, Guillaume Cuypers and Petrus van den Bosch, eds, Acta Sanctorum, June, vol. 5 (Paris: Victor Palmé, 1867), 443C–D; cf. Stevenson in Asser, De rebus gestis Aelfredi, 217–18, and Keynes and Lapidge, trans., in Asser, De rebus gestis Aelfredi, 238). It bears noting here, however, that Eutropius (and thus Paul the Deacon and Landolf Sagax) describes Trajan’s foundation of a province across the Danube after his conquest of Dacia: ‘Daciam Decibalo victo subegit, provincia trans Danubium facta …’ (After defeating Decebalus he conquered Dacia and founded a province beyond the Danube). Eutropius, Breviarium ab urbe condita 8.2.2, ed. Carlo Santini (Leipzig: Teubner, 1979), 50; cf. Henry of Huntingdon, Historia Anglorum 1.25, ed. and trans. Diana Greenway (Oxford: Clarendon Press, 1996), 44–5. If this corruption does in fact date to the late tenth or early eleventh century, then this Danubian origin of the great pagan army offers a very interesting context for the near contemporary elision of Dacia and Dania by Dudo of Saint-Quentin (see p. 000 below).
This ambiguity around the north created broader uncertainties about the relationship of north and west in a European context. While ambiguity around the north itself has been addressed well in terms of continental geography, the ambiguous relationship of north and west becomes clearer when we look to insular geography. Britain has long been associated with a broader notion of the islands of the north-western ocean. For the ancients, the ocean served as a significant point of distinction, between chaos and order, and between the world and the boundless beyond.142Patrick Gautier Dalché, ‘Comment penser l’Océan? Modes de connaissances des fines orbis terrarum du nord-ouest (de l’Antiquité au XIIIe siècle)’, in L’Europe et l’Océan au Moyen Age: Contribution à l’Histoire de la Navigation (Nantes: Cid, 1988), 217–20; e.g., Plato, Statesman 273d–e (trans. C. J. Rowe, in Plato: Complete Works, ed. John M. Cooper (Indianapolis: Hackett, 1997), 315). The islands of the ocean were therefore viewed as liminal spaces beyond the bounds of the world and, by the late republic, it was a common topos that Britain represented a limit of the world. Vergil sets it alongside Libya, Scythia and Oaxes in his first Eclogue, describing its inhabitants as ‘cut off from the whole world’.143Vergil, Eclogae 1.66, ed. Silvia Ottaviano (Berlin: De Gruyter, 2013), 40, trans. Guy Lee (Harmondsworth: Penguin, 1984), 32–3: ‘et penitus toto divisos orbe Britannos’; cf. Horace, Odes 1.35.29–30, ed. D. R. Shackleton Bailey (Leipzig: Teubner, 1995), 38. See also Robert Coleman, Vergil Eclogues (Cambridge: Cambridge University Press, 1977), 86. This topos is exemplified in Solinus’s famous description of Britain as ‘almost another world’, a phrase repeated by Servius to interpret Vergil’s expression.144Solinus, Collectanea rerum memorabilium 22.1, ed. Mommsen, 99–100: ‘Finis erat orbis ora Gallici litoris, nisi Brittania insula non qualibet amplitudine nomen paene orbis alterius mereretur’; Servius, In Vergili Bucolica, 1.66, ed. Thilo, 3:15. See also R. M. Thomson and M. Winterbottom, William of Malmesbury: Gesta Regum Anglorum, vol. 2, General Introduction and Commentary (Oxford: Clarendon Press, 1999), 48–9. For the use of this trope in relation to islands more generally, see Gautier Dalché, ‘Comment penser l’Océan’.
Most Roman geographers situated Britain between north and west, opposed by multiple regions of Europe, as the first in a chain of islands extending to Thule.145Pliny, Naturalis historia 4.16.102–4, ed. Rackham, 2:196–9; Pomponius Mela, De chorographia 3.49–58, ed. Silberman, 81–3; cf. Strabo, Geography 4.5, trans. Jones, 2:253–63. On the early history of Thule, see Kedwards, Medieval Iceland, 129–32. This is also what we find with Martianus Capella and Bede.146Martianus Capella, De nuptiis 6.666, ed. Willis, 235; Bede, Historia ecclesiastica 1.1, ed. Colgrave and Mynors, 14–15. Likewise for Isidore, although he does not provide specific cardinal referents for Britain, the subsequent islands lie on a north-western trajectory and Ultima Thule is found ‘in the northwestern region, beyond Britannia’.147Isi., Etym. 14.6.2–6, trans. Stephen A. Barney et al. (Cambridge: Cambridge University Press, 2006), 294: ‘Thyle ultima insula Oceani inter septentrionalem et occidentalem plagam ultra Brittaniam’. Orosius, however, provides a slightly different picture. He sets Britain on a more directly northern trajectory, explaining that it extends to the north, with Gaul to its south. This is followed, as usual, by the Orkneys and Thule, with the final section on Europe devoted to Ireland, itself running southwest to northeast.148Orosius, 1.2.76–80, ed. Zangemeister, 12. Finally, a few authors of late antiquity situate Britain to the west. Pseudo-Aethicus, of course, places the Atlantic Archipelago within the western ocean, and although there is no island of Thule, he does list a ‘mare Tille’ among the seas, which is presumably a corruption of the ‘mare Thyle’ in Julius Honorius.149Ps.-Aeth., 1.16, ed. Riese, 78; Jul. Hon., Cosm. 14, ed. Riese, 33: ‘Sea of Thule’. Drawing on Julius Honorius, Jordanes moves Thule to the far west of the ocean as a means to emphasise the northern nature of Scanza, the island home of the Goths.150Jordanes, Getica 1.8, ed. Theodor Mommsen, MGH AA 5.1 (Berlin: Weidmann, 1882), 55–6, trans. Peter Van Nuffelen and Lieve Van Hoof (Liverpool: Liverpool University Press, 2020), 224–5: ‘Habet et in ultimo plagae occidentalis aliam insulam nomine Thyle …’ ([The ocean] holds also, in its most western region, another island called Thule …). For an overview of recent literature on the relationship of Thule and Scanza in Jordanes, see Van Nuffelen and Van Hoof, trans., in Jordanes, Getica, 44–8, esp. 48n225. On Jordanes’s use of Julius Honorius, see Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 163, 188. He leaves Britain, however, in its traditional location between Spain, Gaul and Germania without further reference to the cardinal points.151Jordanes, Getica 2.10, MGH AA 5.1, 56. More enigmatic still is Gildas, who in his brief geographical introduction describes Britain as ‘at the end of the world, towards the west or north-west’, but there is no obvious source for this western location.152Gildas, De exidio 3.1, ed. and trans. Michael Winterbottom (London: Phillimore, 1978), 89, 16: ‘Brittannia insula in extremo ferme orbis limite circium occidentemque’. Cf. John Morris in Gildas, De exidio, 148. So while Britain is clearly situated in the north-west, there is no agreement among late ancient or early medieval sources about whether one direction predominates.
By the twelfth century we find authors increasingly emphasising either north or west, often as a way to associate or dissociate islands in the Atlantic with continental Europe. Hugh’s geography, for example, is disproportionately influenced by Orosius, both in the Descriptio and Chronica. While he draws a sharper distinction between continental and insular geography than Orosius, the influence is nevertheless clear.153HoSV, Desc. prol., ed. Gautier Dalché, 133. The first account in the Descriptio is drawn directly from Orosius, with Britain extending to the north from Gaul.154HoSV, Desc. 26, ed. Gautier Dalché, 159. Cf. Orosius, 1.2.76, ed. Zangemeister, 12. Hugh provides a second account of the location of Britain in an extended description of Gaul and Italy near the end of the continental geography, which he has drawn from Hugh of Fleury. This description of the boundaries of Gallia Belgica also follows Orosius except on one point: it omits Britain as a boundary, leaving only the British ocean to the north-west.155HoSV, Desc. 25, ed. Gautier Dalché, 157; Hugh of Fleury, Historia ecclesiastica 5.prol., ed. Leendert Martin de Ruiter (PhD diss., Rijksuniersiteit Gronigen, 2016), 103; cf. Orosius, 1.2.63, ed. Zangemeister, 11, trans. Fear, 43: ‘Gallia Belgica habet … a circio oceanum Britannicum, a septentrione Britanniam insulam’ (Gallia Belgica has … to the north-west, the British Ocean; and to the north, the island of Britain). The italicised section has been removed in the Hughs’ description. We may compare this with the geography in his Chronica, where Britain is listed second to last on the list of islands ‘in the northern ocean from the east to the west’.156Paris, Bibliothèque nationale de France, Latin 15009, 17r: ‘In oceano septemtrionali ab oriente ad occidentem’. Thus in both cases, northernness is given precedence over westernness in the description of Britain.
Lambert also provides a fairly conservative account. Like Isidore, he situates Britain within the list of islands of the ocean, rather than provinces of the world.157On the different ordering of islands in Lambert and Isidore, see Gautier Dalché, ‘Maps in Words’, 229; Patrick Gautier Dalché, ‘De la gloss à la contemplation: Place et fonction de la carte dans le manuscrit du haut Moyen Age’, in Géographie et culture, 746–8; and Derolez, The Autograph Manuscript, 74–5. It is described, following the late antique authorities (with notable cartographic influence), as ‘in the ocean across from Gaul’, with Thule ‘beyond Britain’, and Hibernia and Scotia ‘near Britain’.158For the cartographical connection, see Gautier Dalché, ‘Maps in Words’, 228–9; Lamb., Lib. flor., 51v: ‘Brittannia contra galliam in oceano … Thile ultra brittanniam … Hibernia proxima brittannie’. In this case there is a ‘Hiberus’ added ‘between Spain and Britain’, with the Orkneys set adjacent to Britain on the side of Gothia.159Lamb., Lib. flor., 51v: ‘Hiberus inter hispaniam et brittaniam … Orcades XXXIII inter britanniam et gothiam’. These generally correspond with the placement of islands on his world and European maps, with Britain directly across from Gaul on both. Although this representation of Britain remains traditional, the same cannot be said for the placement of Thule on his maps, where it is found in the far west of the ocean, beyond the south-western edge of Spain.160Lamb., Lib. flor., 241r; Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf. 1 Gud. lat., 69v–70r (Gautier Dalché, MM, Wolf13). Given close association of Lambert’s Europe map with different origin stories for the Franks, including the narrative of Freculf of Lisieux that sees them emerge from Scanzia, it is plausible that Jordanes lies in the background here, although no direct reference to either Thule or Jordanes is contained in the surrounding material.161Lamb., Lib. flor., 234r–241r; on the contents of this section and its complicated history of compilation, see Derolez, The Autograph Manuscript, 164–6. If my suggestion about the map’s relevance to Frankish origins is correct, then this might explain its unusual placement highlighted by Derolez (The Autograph Manuscript, 152). More broadly, though, Lambert and Hugh have both, in their own way, distinguished continental geography from the islands of the ocean.
There is a stronger tendency among some twelfth-century authors to emphasise the westernness of Britain. Honorius provides the clearest example of this trend. Following Orosius, Britain is the last region in his section on Europe, separating it from the later general discussion of islands. This is especially relevant as Honorius specifically notes that some of those later islands (e.g. Avidos) are ‘in Europe’.162Hon., Im. mun. 1.29, 33–5, ed. Flint, 62–6: ‘in Europa’. But what is particularly strange about Honorius’s description is that the islands around Britain likewise follow a westward trajectory, being introduced as ‘islands in the ocean facing the west of Spain’ and the only hint of their northernness comes with the description of Thule: ‘after this to the north the sea is frozen and there is perpetual cold’.163Hon., Im. mun. 1.29, ed. Flint, 62: ‘Contra Hispaniam versus occasum sunt in oceano hee insule … Ultra hanc versus aquilonem est mare congelatum et frigus perpetuum’. There is little solid textual basis for such a western location. It is unclear whether Honorius had read Pseudo-Aethicus, and could have thus drawn on his western ocean.164See Flint in Hon., Im. mun. 46–7, passim, and p. 000 above. Neither is there any good reason to think that Honorius had read Gildas, who could also provide some precedent.165Gildas, De exidio 3.1, ed. Winterbottom, 89.
A more fruitful comparison can be found in cartographical representations. Britain is normally situated opposite France and the Iberian Peninsula often bends up around France towards it, as on the Ebstorf, Hereford, Cotton and Sawley world maps. All these maps represent Ireland behind Britain to the northwest, but this is not the only arrangement. Other maps situate Ireland adjacent to Britain, wrapping further around the continent, normally anticlockwise to the west. In these cases, Ireland is portrayed almost directly across from the north of Spain, as on the Munich map and the Psalter map.166Ireland also sometimes depicted clockwise to the north, as on Lambert’s maps. Honorius may therefore be drawing on a cartographical trend to situate Britain near the western end of Europe, especially where the subsequent islands extend westward around Spain. When taken in combination with the rigorous east-west organisation of his geographical description, which will be addressed in detail later, Britain’s location as the western limit of Europe, rather than the north, is somewhat more understandable.
Honorius is not the only author who situates Britain to the west of the European continent. Guido of Pisa takes a similar line in his discussion of the four sides of Europe at the end of his geography, based closely on the Cosmographia of the Ravenna Anonymous. The largest section by far addresses the north, but it ends with a discussion of the western edge of Europe, focused on Spain and the Britannic ocean:
Spain has as a boundary on one side the Pyrenees mountains next to the shore of the ocean, whereas Spano-Gascony adjoins the Gallic Mediterranean, as well as the same Pyrenees from the other side and the province Septimania. Likewise, from the second side Spain has as a boundary, often called the Gallic-Balearic sea. From the third side, that is from the western side which was noted before, the British ocean. Finally, from the fourth side it has as a boundary the previously noted strait of Gibraltar, which departs from the Gallic sea and enters into the aforesaid western British ocean, and which separates the homeland of the Moors of Africa and the same Spain.167Guido of Pisa, Geographica 130, ed. Schnetz, 142: ‘Hispania habet fines ab uno latere iuxta litus occeani saltum Pireneum, Spanosguasconiam vero iuxta mare magnum Gallicum, ipsumque ex alio latere saltum Pirineum et provinciam Septimanam. a secundo quoque latere habet fines Hispania saepe dictum mare Gallicum Valeriacum. a tertio autem id est ab occidentali quod praediximus mare occeano Britannicum. a quarto demum latere habet fines praemissum fretum Gaditanum, quod exit de mari Gallico et ingreditur in praedictum occeanum occidentalem Britannicum, quodque dividit inter Africae Maurorum patriam et eandem Hispaniam’. Cf. Ravennatis anonymi cosmographia 4.45–6, ed. Schnetz, 82–3. I have identified the rather puzzling ‘Valeriacum’ according to its use by the Ravenna Anonymous, who Guido follows here. That work confuses a variety of letters in classical orthography, such as v and b, making the ‘Baliares’ of Pliny ‘Valeares’, which are then situated in the ‘mare Gallicum Valeriacum’, itself presumably a derivative of Pliny’s classical ‘mare Baliaricum’ (see Louis Dillemann, La Cosmographie du Ravennate (Brussels: Latomus, 1997), 19, 191–2). Cf. Pliny, Naturalis historia 3.5.74–8, ed. Rackham, 2:56–7, and Ravennatis anonymi cosmographia 5.27, ed. Schnetz, 102–3.
Though Guido does not discuss Britain beyond his references to the British ocean, his major source, the Ravenna Anonymous, does. It also situates Britain in the western ocean, and the puzzling description of Gaul lying to its west and Thule to its east, may suggest that this chain of islands is set on more of a western trajectory than a northern one.168Ravennatis anonymi cosmographia 5.31–2, ed. Schnetz, 108; see also Dillemann, La Cosmographie, 195–6. For these authors the tradition of Julius Honorius remains a significant influence, particularly given their organisational use of oceans.
Turning back to the cartography, we can observe how the fluctuation between Spain and Gaul becomes entangled with the representation of Britain’s integration into Europe in the west or its separation as part of an insular world in the north. A particularly stark example of northern displacement can be found on the highly idiosyncratic map in the Thorney Abbey Computus.169Oxford, St. John’s College, MS 17, 6r; Gautier Dalché, MM, Oxf23. On this map, see Edson, Mapping Time and Space, 86–95. Britain, Ireland and Thule have wandered so far north here that they lie east of the northern axis, adjacent to Constantinople and Asia Minor. What is more, the presentation of these islands is so liminal that they break the boundaries of the map. While Britain is still situated within the ring of the ocean, Ireland and Thule extend beyond it and lie outside the outer margin of the map itself. Conversely, we can find a striking example of western displacement on the early thirteenth-century, Icelandic Larger Viðey Map.170Reykjavík, Stofnun Árna Magnússonar, GkS 1812 III 4to, 5v–6r; reproduction, transcription and commentary in Kedwards, Medieval Iceland, 197–229, on this map: 101–18. Here we find Scotia, Anglia and Ibernia on the central east-west axis of the map, in line with the list of continental European provinces, sandwiched between Gaul and Spain. As Dale Kedwards notes, the placement of Britain may in part be explained by the grouping of Normandy, Brittany and Gascony with Spain, thus allowing the three islands to demarcate the Angevin continental holdings (until 1204) from the kingdom of France.171Kedwards, Medieval Iceland, 112. It may also be relevant, though, that the movement of Britain into the continent has left the north-western region of the map free for Iceland and Thule, which can now be situated immediately underneath the label ‘Europa’. Alfred Hiatt identifies a similar dynamic in the contrast between the Cotton world map (fig. 3) and the geography of Gerald of Wales, which may be viewed in light of a broader shift from Scandinavian to Norman and French connections.172Alfred Hiatt, ‘From Pliny to Brexit: Spatial Representation of the British Isles’, Postmedieval: A Journal of Medieval Cultural Studies 7, no. 4 (2016): 516–19. Gerald frames Britain and Ireland as islands of the western ocean and his regional maps present a striking integrated vision of Britain’s connection to the continent.173Dublin, National Library of Ireland, MS 700, 48r; Gerald of Wales, Topographia Hibernica 1, ed. John J. O’Meara, ‘Giraldus Cambrensis in Topographia Hibernie. “Text of the First Recension”’, Proceedings of the Royal Irish Academy: Archaeology, Culture, History, Literature 52 (1948–50): 120: ‘Insularum occidentalium hec ultima [sc. Hibernia] … habet; Britanniam maiorem ab oriente; solum occeanum ab occidente’ (This last of the western islands [sc. Ireland] has Great Britain to the east and only ocean to the west). See also Thomas O’Loughlin, ‘An Early Thirteenth-Century Map in Dublin: A Window into the World of Giraldus Cambriensis’, Imago Mundi 51, no. 1 (1999): 24–39. So also, it may not be incidental that, unlike Hugh and Lambert, Honorius is the only insular author in the group, an author who, in his connection with the school of Anselm of Canterbury, was deeply involved in the Norman integration of England into continental networks. Finally, where the geographies of Hugh and Lambert are associated with maps that provide a highly schematic presentation of Britain as an undifferentiated island of the ocean, mappae mundi from England (e.g. the Sawley or Hereford maps), by contrast, more often embed their representations of Britain physically within the European continent.
* * *
These case studies illustrate the contradictions that had developed out of the classical geographical tradition by the twelfth century. Although the geography itself often remained largely the same, its underlying conceptual foundations had shifted. The classical concerns for river boundaries and relative location, which had already been problematised in patristic geography and late antique textbooks, were increasingly giving way to the use of cardinal points as autonomous and absolute geographical regions of the world. Twelfth-century authors were increasingly conceptualising the world through the combination of these four regions alongside the traditional three parts. This synthesis was driven both by the use of Pseudo-Aethicus alongside the traditional late Roman geographies of Orosius and Isidore as well as the increasing use of cartography in the conceptualisation of geographical space. The resultant difficulties become most evident at the boundaries of the known world. The south developed into an increasingly autonomous region, comprised centrally of Egypt, Ethiopia and Africa, and the cartographically distinct Red Sea gained corresponding prominence as its border. We even find Egypt being grouped for the first time with Africa rather than Asia. Likewise, the north-western fringes of Europe grew more ambiguous, especially as authors needed to expand on their frequently sparse late antique sources. The Atlantic Archipelago serves to highlight this ambiguity, as authors increasingly provide radically different accounts of its location vis-a-vis the European continent. Indeed, as the north was increasingly integrated into Europe politically and geographically, Britain vacillated between its traditionally liminal position in the north and its integration into a western Europe.
Reading the World
As I have emphasised throughout the preceding section, the cardinal points reveal how the inconsistency between different geographical models created structural problems for medieval authors, as they sought to harmonise or simply even make sense of the received geographical tradition. We can approach this same collection of problems, though, from a different direction and in this section I want to think especially about how medieval authors structured their geographies and what that means for the way that medieval readers engaged with them. This poses some immediate problems. Above all, we obviously cannot inhabit the mind of a twelfth-century reader and any attempt to reconstruct the thought processes or mental associations that such a reader might have had as they moved through a text is fraught with difficulties. This problem, though not unique, is particularly acute for the history of medieval geography, where particularly in the anglophone tradition, there has been a tendency to focus centrally on maps.174Gautier Dalché, ‘Introduction’, 8–11; Gautier Dalché, ‘Considérations intempestives’. Though maps offer an evocative image of how medievals viewed the world, they are problematic to ‘read’ and therefore there is a tendency to impose pre-conceived interpretations.175Stephen McKenzie, ‘The Westward Progression of History on Medieval Mappaemundi’, in Harvey, ed., Hereford World Map, 335–44, highlights many of the issues around ‘reading’ maps. Gautier Dalché, ‘Introduction’, 10, also notes some of the more egregious examples. The study of maps has undoubtedly spurred a more synthetic understanding of the way in which medieval authors carefully created connections between history, geography and theology, but without the guidance of closely related texts it is difficult to draw conclusions that are not either highly general or overly dependent on speculative readings of particular motifs. What I want to do here, therefore, is think about how particular textual structures and formulae can guide our understanding of medieval ‘reading’ practices when considering the way in which medieval authors parsed world geography. Beginning from these textual formulae I will address how the ordering of the three parts developed in concert with the use of maps into the twelfth century and then move on to some of the broader ordering principles of early twelfth-century geographical texts.
Ordering the Three Parts
As we have already seen, the tripartition of the oecumene was a fundamental structure for geographical knowledge in the Middle Ages. This was embodied in the formula that opens most geographical texts, explaining that the world is divided into three parts: Asia, Europe and Africa, and that Asia is as large as the other two combined. Despite its ubiquity there has been relatively little discussion among modern scholars of the actual formulae that are used.176See Mauntel, Erdteile, 257–68. See also Natalia I. Petrovskaia, ‘La formulation tripartite du monde dans les encyclopédies et textes littéraires du Moyen Âge’, in La Formule au Moyen Âge IV, ed. Élise Louviot, Charles Garcia and Stephen Morrison (Turnhout: Brepols, 2021), 197–217, whose theory that the two orderings (As, E, Af and As, Af, E) can be divided respectively into an Isidorian and Orosian tradition strikes me as unconvincing; on Petrovskaia’s application of the Translatio imperii, cf. Wolever, ‘Translatio imperii’. Oschema, for example, has emphasised its general importance in establishing the topos of Europe as tertia pars mundi, but maintains that the particular wording of the formula is not especially relevant.177Oschema, Bilder von Europa, 206–8. Yet, by the twelfth century, there is divergence both in order of the three parts in the formula itself and correspondingly in the order of geographical exposition. So while Lambert of Saint-Omer and Honorius Augustodunensis follow the traditional, late antique formula both listing and describing the three parts in the order: ‘Asia, Europe, Africa’, with Hugh of Saint Victor there is a change.178Hon., Im. mun. 1.7, ed. Flint, 52; Lamb., Lib. flor., 19r. In the Descriptio, Hugh has swapped Europe and Africa both in his formula and correspondingly with their order in the text itself.179HoSV, Desc. 7, ed. Gautier Dalché, 138–9. Given both the relationship between cartographical and textual representations of space and the cartographical basis of Hugh’s text, this divergence offers an interesting insight into the intersection of map and text, and into the way that medieval authors conceptually structured world geography.180Gautier Dalché, ‘Maps in Words’.
There is a longer history to the development of these formulae. Even though both begin with Asia, this feature only solidified among the Christian authors of late antiquity.181Mauntel, Erdteile, 155–65. Prior to this, there was no one ubiquitous formula like we find among the Fathers.182Mauntel, Erdteile, 101–3; Salway, ‘Putting the World in Order’, 213–16; Ingrid Baumgärtner, ‘Europa in der Kartographie des Mittelalters: Repräsentationen – Grenzen – Paradigmen’, in Baumgärtner and Kugler, eds, Europa im Weltbild, 11–13; Oschema, Bilder von Europa, 89–90. Roman authors tended to give primacy of place to Europe, with ‘Europe, Africa, Asia’ and, more often, ‘Europe, Asia, Africa’ representing the most common orderings.183Libya and Africa have been amalgamated under ‘Africa’. E, Af, As: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Lucan, 6.817, ed. Shackleton Bailey, 162; Isi., Etym. 13.16.1. E, As, Af: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Cicero, De natura deorum 2.165, ed. W. Ax (Berlin: De Gruyter, 2008), 116; Strabo, Geography 1.4.7, trans. Jones, 1:245; cf. Martianus Capella, De nuptiis 6.622, ed. Willis, 217. Some Christian authors also used these orderings, e.g. Boethius, De diuisione, PL 64, 888C; Cassiodorus, Historia tripartita 3.3.13, ed. W. Jacob and M. Adriaen, CSEL 71 (Vienna: Tempsky, 1952), 141; and Isidore of Seville, De natura rerum 48.2, ed. Fontaine, 324–5. Finally, these orderings were transmitted via Pliny to some early medieval authors: Bede, De natura rerum 51, CCSL 123A, 233–4; Bede, Expositio actuum apostolorum 10.7, ed. M. L. W. Laistner, CCSL 121 (Turnhout: Brepols, 1983), 49; and Dicuil, Liber de mensura orbis terrae prol.6, 1–3, ed. Tierney, 44–55. It is with the major Latin Fathers that we see the solidification of a new and consistent formula: ‘Asia, Europe, Africa’.184Augustine, De civitate Dei 16.17, CCSL 48, 521; Orosius, 1.2.1–11, 6.17.4, ed. Zangemeister, 5, 218; Quoduultdeus, Liber promissionum, Dimidium temporis 11.19, ed. R. Braun, CCSL 60 (Turnhout: Brepols, 1976), 204; Jerome, In Hiezechielem 2.5.5–6, CCSL 75, 56; Jerome, In Dan. 1.4.7b–8, CCSL 75A, 813; Jerome, In epistulas Paulinas ad Titum 1.12–14, ed. F. Bucchi, CCSL 77C (Turnhout: Brepols, 2003), 30; Cassiodorus, Institutiones 2.3.10, ed. Mynors, 114; Isi., Etym. 14.2.1; Ps.-Aeth., 1.1, ed. Riese, 71; Jordanes, Getica 1.4, MGH AA 5.1, 54; Bede, Homeliae evangelii 2.13, ed. D. Hurst, CCSL 122 (Turnhout: Brepols, 1955), 269. It is worth noting the prominence of Sallust, Bellum Iugurthinum 17.3, ed. Kurfess, 68, which lists Asia and Europe in that order after noting that Africa is normally thought of as the world’s ‘third part’. It should be unsurprising therefore that this is the formula preferred by most early medieval authors as, besides Bede, all the important Christian authorities, above all Orosius and Isidore, use it more or less consistently.185There is one off-hand occurrence of an As, Af, E ordering in Orosius, 6.1.6, ed. Zangemeister, 187; cf. Petrovskaia, ‘La formulation tripartite’, 206.
Early medieval geography remains understudied, with only a handful of monographs covering a wide variety of material. As such, it is difficult to offer concrete general remarks about the state of early medieval geography and how it was taught.186See generally Lozovsky, Earth Is Our Book. There is somewhat more for late antiquity, e.g. Merrills, History and Geography, and Gautier Dalché, ‘L’enseignement de la géographie’. Gautier Dalché has pointed to some tantalising evidence about the use of maps in education in late antiquity and the early Middle Ages, but Lozovsky has more generally emphasised the fundamental textual conception of the world in the late Carolingian geographical tradition.187Gautier Dalché, ‘L’enseignement de la géographie’, 163–74, and Patrick Gautier Dalché, ‘Cartes et enseignement de la géographie durant le haut Moyen Âge: L’exemple d’un manuel inédit’, in Du copiste au collectionneur. Mélanges d’histoire des textes et des bibliothèques en l’honneur d’André Vernet, ed. Donatella Nebbiai-Dalla Guarda and Jean-François Genest (Turnhout: Brepols, 1999), 54; Lozovsky, Earth Is Our Book, ch. 4, esp. 138. In keeping with Lozovsky’s observation we predominantly find the fundamentally textual formula of late antiquity, ‘Asia, Europe, Africa’, in these early medieval sources.188Situs orbis 5.1, ed. Gautier Dalché, 164; Descriptio terrarum, Albi, Bibliothèque municipale, MS 29, 58v; Versus de Asia et de uniuersi mundi rota, ed. Glorie, CCSL 175, 441–54; Commentaria Bernensia 9.411, ed. Hermann Usener, Scholia in Lucani Bellum civile: Pars prior (Leipzig: Teubner, 1869), 301; Rabanus Maurus, De universo 12.2, PL 111, 333C–D.
Hugh’s deviance from the traditional formula is hardly isolated, but fits into a broader trend toward ‘Asia, Africa, Europe’ in the central Middle Ages. In the early twelfth century, it appears to be especially prominent among authors associated with the schools of northern France. Thus, for example, although he refers his reader to Orosius’s geography, Otto of Freising presents the three parts in this new order.189Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 37; cf. William of Conches, Dragmaticon 6.5, CCCM 152, 193; William of Conches, Glossulae super Boethium 3.m12, ed. L. Nauta, CCCM 158 (Turnhout: Brepols, 1999), 208; Sigebert of Gembloux, Chronica 738, ed. Ludwig Bethmann, MGH SS 6 (Hannover: Hahn, 1844), 331 [hereafter Sig., Chron.]. But the trend has deeper roots. We find notable examples of this new formula already in the eleventh century. The Cotton world map, for example, accompanies a copy of Priscian’s Periegesis, whose incipit links it to the facing map, explaining that a ‘suitable map’ has been added ‘to this work about the three parts, namely Asia, Africa and Europe’.190London, British Library, Cotton MS Tiberius B V, 57r, 60v: ‘huic operi de tribus partibus, videlicet Asia, Africa, Europa, mappam depinxerat aptam’; transcription in Gautier Dalché, ‘L’héritage antique’, 46n41. On the relationship of the incipit to the map, see Alfred Hiatt, ‘The Cotton Mappa mundi as analogue to the Albi world map’, in Vagnon and Victor, eds, La Mappa mundi d’Albi, 199–201. This is in spite of Priscian’s own presentation of Africa, Europe and then Asia.191Priscian, Periegesis 15, ed. van de Woestijne, 38. There are fewer examples prior to the eleventh century. A few cases can be found in relationship with Noah’s sons, for which Asia, Africa, Europe is the natural order.192Gn 5:31, 6:10, 7:13, 9:18, 10:1; Eucher of Lyon, Instructionum ad Salonium 1.19, ed. C. Mandolfo, CCSL 66 (Turnhout: Brepols, 2003), 87; Bede, In principium Genesis 3.10.1–2, ed. C. W. Jones, CCSL 118A (Turnhout: Brepols, 1967), 142. Both the Liber de mundi institucione, ed. Michel Zimmerman, ‘Le monde d’un catalan au Xe siècle: Analyse d’une compilation isidorienne’, in Le métier d’historien au moyen-âge: Etudes sur l’historiographie médiévale, ed. Bernard Guenée (Paris: Publications de la Sorbonne, 1977), 74–8, and Rabanus Maurus, De universo 2.1, PL 111, 34D associate the three parts with Noah’s sons, but they nevertheless follow the As, E, Af ordering in their geographical description. More significantly, there is fragmentary evidence of this development in the Carolingian schools of northern France. For instance, in a ninth-century version of the Anonymous Cantabriensis commentary on Martianus Capella, which Lozovsky suggests may originate in Soissons, the glossator deviates from the other British manuscripts in this family and lists the three parts: ‘Asia, Africa, Europe’.193Paris, Bibliothèque nationale de France, Latin 8669, 77r; transcription in Lozovsky, Earth Is Our Book, 134n109, cf. 116. Cf. Heiric of Auxerre, Homiliae per circulum anni, pars aestiua 9, ed. R. Quardi, CCCM 116B (Turnhout: Brepols, 1994), 71; Glossae collectae e primo secundoque libris Orosii historiarum contra paganos 1.2.95, ed. Olivier Szernwiniack, ‘Un commentaire hiberno-latin des deux premiers livres d’Orose, Histoires contre les paines’, Archivum latinitatis medii aevi 51 (1992/3): 70–1, where the comment on the three parts is a deviation from the section of Isidore being copied. A French origin would also tally well with Richer of Saint-Remi, Historiarum libri IIII 1.1, ed. Hartmut Hoffmann, MGH SS 38 (Hanover: Hahn, 2000), 36, but examples of an As, Af, E ordering are already found in Italy by the mid-eleventh century (e.g. Bruno of Segni, Commentaria in Ioannem 3.54, PL 165, 599B; cf. Mauntel, Erdteile, 261n1118) and in Salzburg by the twelfth century, where the first scribe of a Waltharius manuscript has written ‘Asia, Affica, Europa’ twice as penmanship practice on the preceding folio, perhaps following the poem’s opening line: ‘Tertia pars orbis, fratres, Europa vocatur’ (The third part of the world, brothers, is called Europe). Vienna, Österreichische Nationalbibliothek, Cod. 289, 103r; Waltharius 1, ed. Karl Strecker, MGH Poetae 6.1 (Weimar: Hermann Böhlaus, 1951), 24.
Returning to the early twelfth century, this development is not only associated with northern France, but is disproportionately associated with cartography.194The Expositio mappe mundi by Roger of Howden (?), ed. Patrick Gautier Dalché, Du Yorkshire a l’Inde: Un ‘géographie’ urbaine et maritime de la fin du xiie siècle (Roger de Howden?) (Geneva: Droz, 2005), 143–64, also follows the order As, Af, E. See generally, Gautier Dalché, ‘Comment et pourquoi décrire une mappemonde’. This opens up the interesting possibility that the formula has a principled relationship with the use of maps, especially in the schools of northern France. If this is indeed a product of the clockwise reading of a map, it would tally well with the evidence we have to suggest that maps were gaining increasing currency in geographical education and in shaping geographical thought from the early twelfth century.195Gautier Dalché, ‘Maps in Words’, 227–34; Terkla, ‘Hugh of Saint-Victor’. Hugh of Saint Victor’s Chronicle is an instructive example. Even though it begins by listing the three parts of the world: Asia, Europa, Africa, the actual regions themselves are described in the ‘cartographic’ ordering: Asia, Africa, Europa.196Paris, Bibliothèque nationale de France, Latin 15009, 16r–v. So while the textual formula is used, Hugh’s actual geography is organised according to a clockwise movement around a circular map.
Ordering Geographical Texts
Naturally we should wish to ask then why Honorius and Lambert do not follow the same ordering, as both are also identified by Gautier Dalché as authors whose geographies are cartographically influenced. To address this, we will need to think more carefully about the organisation of these texts. In the case of Lambert, the first major geographical section (chapter three, beginning folio 19r) is accompanied by a T-O diagram containing the names of the peoples in each part of the world. It is tempting to suggest that, due to the highly schematic nature of the diagram and its heavy textual content, the formula pertains also to the way in which Lambert intends the map to be read: left to right across Asia, then across Europe and Africa. But as with most other things, the organisation of geography in the Liber Floridus is eclectic. His lists of the provinces, peoples and cities of the world follow roughly an Asia, Europe, Africa ordering, starting respectively with Assyria, the Indians and, excluding Rome, the cities of India; likewise, each reaches Europe roughly two-thirds of the way through and ends in Africa with Ethiopia for the former two and Libya for the latter.197Lamb., Lib. flor., 48r–50r. The list of islands, on the other hand, runs clockwise, beginning in the east.198Lamb., Lib. flor., 51v. Finally, the rivers of the world are divided up between east, west, north and south, in that order.199Lamb., Lib. flor., 52v. Unsurprisingly, these orderings generally reflect the organisation of his main sources: Isidore and Pseudo-Aethicus.200Isi., Etym. 14.3–5; Ps.-Aeth., 1.7–12, 20–5, 33–7, ed. Riese, 74–7, 81–8, 86–7; cf. Gautier Dalché, ‘Maps in Words’, 228–9; Derolez, The Autograph Manuscript, 67, 73–5. The list of the islands, however, cannot be wholly explained by Isidore, who begins around Britain and generally progresses anticlockwise.201Isi., Etym. 14.6.2–13. Derolez suggests the source of the islands is simply Isidore, though Gautier Dalché highlights various differences between Lambert and Isidore, which he plausibly attributes to Lambert’s use of a map to aid his activity of compilation. Derolez, The Autograph Manuscript, 74n108; Gautier Dalché, ‘Maps in Words’, 228–9. While the eclectic nature of this material is explained in part by the compilatory nature of the text, it does seem to suggest that Lambert himself did not have an overarching geographical schema in mind and was therefore happy to followed whatever text he was using. It is only with the islands that we see a new structure, albeit one that does not precisely reflect the placement of islands on his own mappa mundi.202Though the grand mappa mundi from the autograph manuscript has been lost, there is a version from a late twelfth-century copy in Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf. 1 Gud. lat., 69v–70r, which is normally taken to be accurate. Danielle Lecoq, ‘La Mappemonde du Liber Floridus ou La Vision du Monde de Lamber de Saint-Omer’, Imago Mundi 39 (1987): 9–49; Gautier Dalché, ‘De la glose’, 746–7; Gautier Dalché, ‘Maps in Words’, 227–31. Thus Lambert may simply have turned to a map to aid him with the unusually haphazard organisation of the islands in Isidore.203Gautier Dalché, ‘De la glose’, 746–9.
Unlike Lambert and Hugh, Honorius Augustodunensis does not engage so obviously with maps in his Imago mundi, instead presenting the text itself as a sort of mappa mundi.204Gautier Dalché, ‘Maps in Words’, 230–1. He follows Isidore near verbatim in his description of the three parts, with one very subtle alteration. Isidore describes them anticlockwise, starting from the south: ‘Asia extends from the south through the east to the north, but Europe from the north to the west, Africa from the west to the south’.205Isi., Etym. 14.2.2: ‘Nam Asia a meridie per orientem usque ad septentrionem pervenit; Europa vero a septentrione usque ad occidentem; atque inde Africa ab occidente usque ad meridiem’. I have modified the translation of Barney et al., 285. Cf. Augustine, De civitate Dei 16.17, CCSL 48, 521. Honorius, however, begins in the north and reverses the movement: ‘Asia extends from the north through the east to the south, Europe from the west to the north and Africa from the south to the west’.206Hon., Im. mun. 1.7, ed. Flint, 52: ‘Asia a septentrione per orientem usque ad meridiem, Europa ab occidente usque ad septentrionem, Affrica a meridie usque ad occidentem extenditur’. As is clear from the jump from south to west and north to south that punctuates Honorius’s account, his description of the extent of each region has been reversed: the south, east, north, west sweep in Isidore has become north, east, south, west. Thus, despite the traditional ordering of the three parts, a clockwise reading of the world now underlies Honorius’s description. The change is especially striking as Isidore structures his subsequent description of Europe and Africa from north to south, which Honorius follows.207Both begin their discussion of Europe with Scythia in the north-east and end with Spain in the south-west and they begin their discussion of Africa with Libya in the north-east and end with Ethiopia in the south. What is more, Honorius uses the clockwise (Asia, Africa, Europe) ordering in reference to a (likely fictive) mappa mundi in his De animae exsilio et patria.208Honorius Augustodunensis, De exsilio et patria animae, PL 172, 1244D: ‘In hac Aratus mappam mundi expandit, in qua Asiam, Africam, Europam ostendit’ (In this [city of Geometry], Aratus lays out a world map, on which Asia, Africa and Europe are shown …); cf. Gautier Dalché, ‘Maps in Words’, 230n28. So while Honorius’s text is clearly based on textual geographies, as we see in his close adherence to the structure of Isidore’s text, the change to a clockwise reading may suggest an underlying influence of cartography on Honorius’s mental geography.
Beyond this, Honorius has a rigid and consistent mode of geographical organisation within his text, whose schematic and hierarchical nature Gautier Dalché clearly illustrates.209Gautier Dalché, ‘Maps in Words’, 236. There is also a clear geographical logic to the text’s movement through the world. Honorius follows a similarly hierarchical structure which remains faithful to Isidore in its broad strokes, but nevertheless diverges in some significant ways. In the first case there is the broad division between the three parts, following the traditional order. But while Honorius moves through the provinces in roughly the same order as Isidore, their manner of description is notably different. Isidore normally begins with an enumeration of each province’s boundaries to the north, south, east and west. Honorius, on the other hand, follows a more east-west structure. He divides Asia into three slices, for which the northern and southern boundaries are typically noted just once, after which only eastern and western boundaries are provided. So while Paradise is the first region of Asia for both, Isidore simply notes that ‘Paradise is located in the east’, whereas Honorius makes it Asia’s eastern boundary: ‘The first region [of Asia] in the east is Paradise’.210Isi., Etym. 14.3.2, trans. Barney et al., 285: ‘Paradisus est locus in orientis partibus constitutus’; Hon., Im. mun. 1.8, ed. Flint, 52: ‘Huius prima regio in oriente est paradysus’. India, as a result, receives no eastern boundary, but is defined rather by its western boundary, the Indus River, whose northern and southern termini, the Caucus mountains and the Red Sea, define this central ‘slice’ of Asia. India is followed by Parthia, from the Indus to the Tigris; Mesopotamia, from the Tigris to the Euphrates; and Syria, from the Euphrates to the Mediterranean. Honorius makes this organisational structure explicit when he presents these regions as the northern boundary of Egypt, the southern ‘slice’ of Asia, noting that: ‘beginning from the east, these aforementioned regions are extended in a straight line to the Mediterranean’.211Hon., Im. mun. 1.10–17, at 17, ed. Flint, 53–7: ‘Hee superius dicte regiones ab oriente incipientes, recta linea ad Mediterraneum mare extenduntur’. Northern Asia is defined similarly in terms of this central ‘slice’.212Hon., Im. mun. 1.18, ed. Flint, 58: ‘Suprascriptis regionibus versus aquilonem annectuntur he regiones’ (These regions adjoin the previously discussed regions to the north).
This same pattern is apparent in Honorius’s description of Africa, which also emphasises an east-west reading: ‘It arises on the east of the Indus River, and turning through the south, extends into the west’.213Hon., Im. mun. 1.30, ed. Flint, 63: ‘Hec in oriente Indi fluminis surgit, et per meridiem vergens, in occidentem tendit’. No further boundaries are provided for the northern provinces from Libya to Mauritania, just the major cities, and so also when the text turns to Ethiopia ‘toward the south’.214Hon., Im. mun. 1.32, ed. Flint, 64: ‘Versus meridiem vero est Ethiopia …’. The east-west emphasis nevertheless remains with the opposition of a city in the east then one in the west, and ending with a final city ‘on the outer edges of Africa to the west’.215Hon., Im. mun. 1.32, ed. Flint, 64: ‘In extremis finibus Africe versus occidentem …’.
Honorius’s east-west framework is less consistent in Europe. There is still an overall east-west trajectory, but the north-south axis appears more present throughout. Europe begins in the north with the Riphaean mountains, the Tanais river and the Meotide swamps. This northern emphasis pervades this section of Europe, with northern and southern boundaries listed for Scythia, Germania superior and Germania inferior. Similarly, these boundaries are not as standardised as with Asia. For example, Scythia begins ‘from the Tanais river [and] is extended to the Danube on the south’.216Hon., Im. mun. 1.22, ed. Flint, 59: ‘A Tanai fluvio … versus meridiem usque ad Danubium porrigitur’. The south of Europe has a somewhat more consistent east-west focus, beginning with Greece, bounded on the south by the Mediterranean; then Pannonia superior to the Apennines, with the Danube to the north; Italy, from the Alps to the Mediterranean, coupled with a discussion of the northern course of the Rhine from the Alps to the ocean; Gaul, from the Rhine, with the British ocean to the north; Spain, which terminates at the western ocean; and finally Britain, which faces Spain against the west in the ocean, beyond which to the north the sea is frozen and continually cold.217Hon., Im. mun. 1.25–9, ed. Flint, 60–3.
This is the messiest section of Honorius’s geography as he brings his classical sources in line with a more contemporary view of Europe through the integration of a wider north including Denmark and Norway.218Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–73; Fraesdorff, Norden, ch. 3; Klaus Oschema, ‘Der Europa-Begriff im Hoch- und Spätmittelalter. Zwischen geographischem Weltbild und kultureller Konnotation’, Jahrbuch für europäische Geschichte 2 (2001): 196–8. Nevertheless, if we reconsider this section in terms of the spatial logical of a medieval map, it creates two east-west slices divided by the Danube, the traditional north-south boundary.219Cf. Oschema, ‘Europa-Begriff’, 197. At the same time, Honorius’s European geography highlights the importance of the classical north-south dichotomy and the way that twelfth-century authors sought to negotiate it. The need to repeatedly note north and south borders speaks also to the increasing complexity that this process of negotiation created for Europe as compared with Asia, where the Caucuses and the Red Sea, or Africa, could straightforwardly divide the entire region.
This sort of conventional east-west movement in geographical exposition is not restricted to Honorius or the Imago mundi. It is already evident in patristic geography, most obviously with the relocation of Asia to the front of the list.220Mauntel, Erdteile, 141–52, 155–65. But there is a wider trend in the twelfth century. The systematic nature of this convention is no less explicit in Hugh’s Chronica, which structures its geography very similarly to Honorius’s. Asia is similarly divided into three strips and Hugh notes that both the north and south sides of Asia extend ‘from the eastern ocean towards the west’.221Paris, Bibliothèque nationale de France, Latin 15009, 16r (emphasis added): ‘In latere asie ad aquilonem ab occeano orientali occidentem uersus porrigitur / [s]cythia’; ‘In latere australi asie porrigitur mare ruberum ab oceano orientali occidentem uersus usque ad confinia arabie et palestine’ (On the side of Asia to the north, Scythia extends from the eastern ocean towards the west; On the southern side of Asia, the Red Sea extends from the eastern ocean towards the west up to the boundaries of Arabia and Palestine). There are equivalently explicit statements in his description of Africa and Europe.222Paris, Bibliothèque nationale de France, Latin 15009, 16r–v. Naturally this is adopted by Richard of Saint Victor in his Liber exceptionum, which is the most explicit, explaining, for instance, that ‘the regions of Africa extend from east to west thus …’.223Richard of Saint Victor, Liber Exceptionum 1.3.3, cf. 1.3.1, 4, passim, ed. Jean Châtillon (Paris: Vrin, 1958), 124, cf. 122, passim: ‘Affrice regiones ab Oriente ad Occidentem porriguntur sic …’. So while this section offers only a sketch, the conventional nature of this type of east to west movement would repay further research, particularly given its broader resonance in medieval thought, as we see already in Honorius’s metaphor of time as a rope stretched from east to west.224Hon., Im. mun. 2.3, ed. Flint, 92–3; on some of the wider historical resonances, see Wolever, ‘Translatio imperii’.
As the organisation of geographical texts from the early twelfth century shows, despite their ostensible deference to late antique authorities, these authors were engaging with the world in new ways. From the eleventh century, emanating especially from the schools of northern France, we see an increasing number of authors altering the formulae and textual organisation of Isidorian and Orosian geography. By as subtle a change as the reversal of Europe and Africa in a list, these authors reveal a fundamental reversal of the ordering of world geography, from anticlockwise to clockwise. This shift comes hand in hand with the increased use of cartography in the schools and of its influence on textual geographies. The organisation of these texts also shows an increasing interest in the schematisation of world geography. This is already evident in the strictly hierarchical nature of Honorius’s Imago mundi, but the developments are broader than this, with both Honorius and the school of Saint Victor evidencing a more systematic presentation of world geography according to an east-west order of exposition.
Framing the World
The chapter began with a caution about the need to address medieval geography on its own terms. In this context, we have seen how the cardinal points represented an important framework for conceptualising the world in the twelfth century. Medieval authors used them to organise the content of the world, both in the structure of their maps and geographical texts, through the often self-conscious interplay of three- and four-part models or in concert with eight or twelve winds.225Recent scholarship has increasingly highlighted the fundamental plurality of models in medieval geography: Akbari, Idols in the East, 24–50; Baumgärtner, ‘Winds and Continents’; cf. Mauntel, Erdteile, 19, and Nathalie Bouloux, ‘Ordering and Reading the World: The Maps in Lambert of Saint-Omer’s “Liber Floridus”’, in Mauntel, ed., Geography and Religious Knowledge, 97–9. At the same time, this framing is, of course, somewhat artificial. The use of the cardinal points in geography was not independent of cosmology or exegesis. Thus, this last section will draw some of this bigger picture back into focus. Specifically, it will address the inclusion of allegorisations of the cardinal points on Sallust maps and their background in eleventh-century wind diagrams, which will highlight the broader cosmological function of the cardinal points in the central Middle Ages.
Beginning in the second half of the twelfth century, an unusual phenomenon arises in a small group of Italian Sallust manuscripts. On a set of relatively detailed maps included alongside Sallust’s description of Africa in the Bellum Iugurthinum, someone has decided to inscribe standard allegorical interpretations of the cardinal points.226‘Sallust map’ simply refers to any map or T-O diagram included in a manuscript of Sallust’s Bellum Iugurthinum, typically near 17.3, ed. Kurfess, 68. On these maps, see generally Patrick Gautier Dalché, ‘Les cartes des manuscrits de Salluste’, Geographia Antiqua 30 (2021): 177–96; Mauntel, Erdteile, 393–8; Edson, Mapping Time and Space, 18–21; David Woodward, ‘Medieval Mappaemundi’, in Harley and Woodward, eds, History of Cartography, 343–4; and Miller, kleineren, 110–15. On diagrams in classical works, see more broadly Patrick Gautier Dalché, ‘Les diagrammes topographiques dans les manuscrits des classiques latins (Lucain, Solin, Salluste)’, in La tradition vive. Mélanges d’histoire des textes en l’honneur de Louis Holtz, ed. P. Lardet (Turnhout: Brepols, 2003), 291–306. These allegorisations are found on four Sallust maps (fig. 5), all produced in Italy, now housed in Paris (P), Cambridge (C), Milan (M) and Venice (V), and which all share clear similarities both cartographically and in their manuscript context.227Paris, Bibliothèque nationale de France, Latin 6089, 26r; Cambridge, Gonville and Caius College, MS 719, 37v; Milan, Biblioteca Ambrosiana, G 21 Sup., 23v; Venice, Biblioteca Nazionale Marciana, Fondo Ant. Lat. Z 432 (1656), 40r. See respectively Paris: Gautier Dalché, MM, Par15 (C12 2/2, Italy?) and Burger Munk Olsen, Catalogue des manuscrits classiques latins copiés du IXe au XIIe siècle, 5 vols (Paris: CNRS, 1982–2020), 2:343 (C12 2/2, Italy); Cambridge: M. R. James, A Descriptive Catalogue of the Manuscripts in the Library of Gonville and Caius College (Cambridge: Cambridge University Press, 1908), 690–1 (C13, Italy); Milan: Robert Black, ‘The Rise and Fall of the Latin Classics’, Aevum 91, no. 2 (2017): 432 (C12 2/2, Italy; but, 432n147, previously dated C13ex, Venice); Venice: Anna Novokhatko, ‘Eine Liste der Handschriften der im Sallust- und Cicerocorpus überlieferten Invektiven (Sallusti in Ciceronem et invicem Invectivae)’, Eikasmos 13 (2002): 281 (citing Zorzanello’s catalogue C14). The Paris, Cambridge and Milan manuscripts are reproduced in Chekin, Northern Eurasia, 341, 343, 346; likewise, the Paris, Cambridge and Venice manuscripts in Woodward, ‘Mappaemundi’, fig. 18.45, 18.50, 18.74. The maps in C, M and V are evidently closely connected. M and V are nearly identical in layout and content, while C shares the same distinctive division of Africa, the Getulians, Ethiopians and torrid zone as well as the demarcation of Spain, location of Jerusalem and Sidon, and the labelling of mare oceanum among other things.228These features are also found on other Sallust maps, e.g., Paris, Bibliothèque nationale de France, Latin 6253, 52v (Gautier Dalché, MM, Par16). The cartographical relationship between these three and P is much looser with a similar arrangement of Egypt and Jerusalem, but an otherwise largely unrelated depiction of Europe and Africa. What links these four maps is a set of (sloppily copied) inscriptions, drawn ultimately from Rabanus Maurus’s commentary on Genesis, adjoining the four cardinal points: ‘[oriens] lucem fidei et ortum bonorum operum significat’; ‘[meridies] claritatem sapientie et feruorem caritatis significat’; ‘[occidens] uitiorum et peccatorum occasum significat’; ‘[septentrio] constrictionem uoluptatis et mortificationem significat’.229The major variants are East: ortum ] ortorum V M; South: caritatis significat deest C; West: uitiorum ] uiuorum V M, occasum ] occasus C P, significat ] designat M dicitur C; North: constrictionem ] districtionem P, uoluptatis ] uoluntatis V M, et deest V M. ‘the east signifies the light of faith or the rise of good works’; ‘the south signifies the clarity of wisdom and the warmth of love’; ‘the west signifies the fall of vices and sins’; ‘the north signifies the binding of delight and mortification [of desires]’. ‘Of desires’ is missing on the Sallust maps, but found in Rabanus Maurus, Commentariorum in Genesim 3.14, PL 107, 593B–C: ‘Occidens in hoc loco occasum vitiorum et peccatorum designat, Oriens lucem fidei vel ortum bonorum operum, Septentrio constrictionem voluptatis et mortificationem desideriorum. Meridies claritatem sapientiae et fervorem charitatis non inconvenienter ostendit’ (See the above trans. in the order: west, east, north, south). The variants in these readings underscore the close relationship between M and V and may suggest a looser connection between C and P. The inclusion of biblical allegorisations around a world map is unusual, particularly as these are, with the exception of V, purely Sallust manuscripts.230V also contains a selection of Cicero’s orations and a pseudonymous exchange of invectives between Cicero and Sallust. To make sense of these inscriptions, we will need to look to their source in earlier wind diagrams and at the role of the cardinal points in that context.
Content - style instruction: <INSERT FIGURES 5 AND 6 NEAR HERE>
The intermediary source of these allegorisations appears to be a very similar set of allegories inscribed around a collection of wind diagrams from the turn of the eleventh century. Their story begins around 990 at St. Pantaleon in Cologne, where Froumund of Tegernsee was then a monk.231See generally Walter Berschin, ‘Eine griechisch-althochdeutsche-lateinische Windrose von Froumund von Tegernsee im Berlin-Krakauer Codex lat. 4°939’, in Vetustatis amore et studio: Księga pamiątkowa ofiarowana Profesorowi Kazimierzowi Limanowi, ed. Ignacego Lewandowski and Andrzeja Wójcika (Poznań: Uniwersytet Adama Mickiewicza, 1995), 23–30 (repr. in Walter Berschin, Mittellateinische Studien, vol. 1 (Heidelberg: Mattes Verlag, 2005), 285–92). It was there that he copied a version of Boethius’s Consolatio with the commentary of Remigius of Auxerre.232Kraków, Biblioteka Jagiellońska, Nr. 834 (previously Berlin, Staatsbibliothek Prußischer Kulturbesitz, Ms. lat. 4° 939). See Rolf Bermann et al., eds, Katalog der althochdeutschen und altsächsischen Glossenhandschriften (Berlin: De Gruyter, 2005), nr. 45. The database is accessible online at glossen.germ-ling.uni-bamberg.de (accessed September 10, 2023). Besides his subscription, which tells us that the manuscript was written in Cologne, and a collection of miscellaneous material about Cologne and St. Pantaleon, Foumund copied an unusual, multilingual wind-diagram into the beginning of the manuscript.233The subscription is transcribed in Die Tegernseer Briefsammlung (Froumund), ed. Karl Strecker (Berlin: Weidmann, 1925), XI. There are three main elements to this diagram: first, a Greek, Latin and Old High German wind rose, with the first three lines of the poem De uentis from the Latin Anthology written in the central circle;234The Old High German is based on Einhard, Vita Karoli 29, ed. G. H. Pertz, G. Waitz and O. Holder-Egger, MGH SS rer. Germ. 25 (Hannover: Hahn, 1911), 34. Paulo Farmhouse Alberto, ‘The Textual Tradition of the Carmen de uentis (AL 484): Some Preliminary Conclusions with a New Edition’, Aevum 83 (2009): 341–75. second, the letters ADAM have been written alongside the corresponding Greek cardinal directions;235On ADAM and the cardinal points, see, e.g., Augustine, Enarrationes 95.15, CCSL 39, 1352–3, and generally Kochanek, Vorstellung vom Norden, 164–218, esp. 208–13. third and finally, the same allegorisations we find on the maps, albeit drawn more precisely from Rabanus Maurus, have been inscribed around the outside of the wind rose along with a second sentence on the climatic characteristics of each cardinal wind.236Berschin, ‘Windrose’, 28: ‘Oriens lucem fidei vel ortum bonorum operum significans. … – Meridies claritatem sapientie et fervorem caritatis ostendens. … – Occidens vitiorum et peccatorum occasum designans. … – Septentrio constrictionem voluptatis et mortificationem desideriorum exprimit. …’ (see trans. as in Rabanus at p. 000 above). Notable differences between the windrose (WR), Sallust map (TO) and Rabanus (RM) include, Oriens: vel WR RM ] et TO; Meridies: ostendens WR ] ostendit RM significat TO; Occidens: designans WR ] designat RM significat TO; Septentrio: desiderorum WR RM deest TO. While it is very likely that Froumund himself added the second and third elements, there is another earlier copy of the wind diagram itself that could already be found in Cologne attached to a manuscript of Boethius’s Consolatio by the mid-tenth century.237The wind diagram is found in London, British Library, Harley MS 2688, 17r, which was originally part of the first quire of London, British Library, Harley MS 3095. On the dating and location of these manuscripts, see Bernard Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts, 3 vols (Wiesbaden: Harrassowitz, 1998–2014), 2:114, 121, and Sigrid Krämer, Handschriftenerbe des deutschen Mittelalters, 3 vols (Munich: C. H. Beck, 1989–90), 2:417. Bischoff dates the former quire C9 4/4, but expresses some doubts about placing the manuscript in Cologne, suggesting France(?) instead and dates the latter C10 1/2, which Krämer places confidently in the possession of Cologne Cathedral. Froumund’s use of this diagram in his copy of Boethius provides a terminus ante quam around the 990s for the combined manuscript’s presence in Cologne.
The manuscript did not remain in Cologne, but returned with Froumund to Tegernsee in southern Bavaria, where it remained until it was bought by Ludwig of Öttingen-Wallerstein around 1815.238Georg Schepss, Handschriftliche Studien zu Boethius De consolatione philosophiae (Würzburg: Thein, 1881), 3. Three further copies of the diagram were made. The earliest, from Echternach (fig. 5), likely predates Froumund’s departure and may represent part of a broader connection between the Abbey of Echternach and St. Pantaleon in the late tenth and eleventh century.239Paris, Bibliothèque nationale de France, Latin 10195, 1r. See Claudine Moulin and Elvira Glaser, ‘Echternacher Glossenhandschriften’, in Die althochdeutsche und altsächsische Glossographie: Ein Handbuch, ed. Rolf Bergmann and Stefanie Stricker (Berlin: De Gruyter, 2009), 1267–8. Here it has been copied into a more miscellaneous school collection containing centrally Macrobius, Sallust and Calcidius. The diagram itself has been rotated 45° clockwise, the internal text is incomplete and it lacks the ADAM, but nevertheless contains the allegorisations. After Froumond’s return to Tegernsee, another eleventh-century copy was produced in Salzburg, which once again accompanies Boethius, but lacks the allegorisations.240Munich, Bayerische Staatsbibliothek, Clm 15825, 1r. There is a final copy from the middle of the eleventh century, which has escaped the notice of previous scholarship on this diagram and somewhat complicates this otherwise straightforward transmission history. The manuscript was likely produced in the Abbey of Bèze, near Dijon, before 1062, but quickly found its way to the library of Saint-Bénigne in Dijon.241Dijon, Bibliothèque municipale, MS 448, 75r. On this manuscript, see Colette Jeudy and Yves François Riou, Les manuscrits classiques latins des bibliothèques de France, vol. 1 (Paris: CNRS, 1989), 483–500. This wind diagram contains the complete text and allegorisations, but lacks the ADAM and is situated within an eclectic Bedean computus collection. Textual variations in the central De ventis offer a further complication here. All four copies ostensibly connected to Froumund share an otherwise unattested variant of the second line of the poem, whereas the Dijon manuscript contains a variant of the beta text.242The Kraków, London, Paris and Munich manuscripts all have ‘trahuntur’ instead of ‘iugantur’, whereas the Dijon manuscript reads ‘ligantur’. The Dijon manuscript also breaks with the other four and follows the beta text by replacing ‘hos circum’ with ‘hinc quoque sex’; cf. Alberto ‘Carmen de uentis’, 360, 365–9, 374. Nevertheless, the allegorisations in all three manuscripts that contain them represent careful copies of the same text.
Given their presence in southern Germany, it was no great leap for the allegorisations to pass over the Alps into northern Italy over the next two centuries. But we can take one more significant step towards connecting wind diagrams to maps, as the intermediary manuscript – or a close copy of it – likewise still exists. Produced in the twelfth century in the diocese of Augsburg and now housed in the Vatican Library, we find a map ultimately based on Sallust’s Bellum Iugurthinum, but copied alongside a collection of computus material and astrolabe treatises of the eleventh century around which our same allegorisations have been inscribed.243Vatican, Biblioteca Apostolica Vaticana, Pal. lat. 1356, 113r; Gautier Dalché, MM, VatP14, and Ludwig Schuba, Die Quadriviums-Handschriften der Codices Palatini Latini in der Vatikanischen Bibliothek (Wiesbaden: Dr. Ludwig Reichert Verlag, 1992), 34–9. See Gautier Dalché, ‘Les cartes des manuscrits de Salluste’, 184–9. The map itself shares many of the distinctive features of the Milan, Venice and Cambridge Sallust maps (fig. 5), such as the peculiar subdivision of Africa with inter alia vertical stripes for the Ethiopians and the Getuli, the demarcation of Egypt by an unusually eastern Red Sea as well as of Spain from Western Europe and the labelling of the mare oceanum. What is more, this map contains a collection of cosmological diagrams set within and without it, including a zonal diagram in the centre and outer circles corresponding to the ages of man and elements, as well as of course a wind diagram, with the winds clustered around the four cardinal points. It is between the cardinal directions and the outermost circle of the diagram that we find the allegorisations.244Vatican, Biblioteca Apostolica Vaticana, Pal. lat. 1356, 113r; Gautier Dalché, MM, 1:404: ‘Oriens: lucem fidei ortum bonorum operum figurat’; ‘Meridies: claritatem sapientie fervorem caritatis figurat’; ‘Occidens: viciorum peccatorum [occasum] figurat’; ‘Septentrio: constrictionem voluptatis et mortificationem’ (see trans. as in the Sallust maps at p. 000 above). Interestingly, the inscription for the north appears unfinished, lacking both ‘figurat’ (depicts) and the final word of Rabanus’s interpretation: ‘desiderorum’, which is found in the wind diagrams, but is missing in all of the Sallust maps. The placement of this inscription, equally spaced underneath ‘Septentrio’, points to an earlier already damaged or incomplete archetype for both this and our Sallust maps. Finally, given the overwhelming preference for significare in the Sallust maps, it is interesting to see that this map has standardised the term of allegorisation. Indeed, given that ‘figurat’ has been abbreviated ‘fig’, it is not hard to imagine that the seemingly hurried scribes of the Sallust manuscripts mistook the f for a long-s and read ‘significat’ or indeed that the scribe of this manuscript, working from a damaged exemplar, mistook a long-s in ‘sig’ for f. Whatever the case, this map undoubtedly represents a key intermediary stage between the wind diagrams and Sallust maps.
As interesting as this transmission history is, it still leaves us with the question of why. In a general sense, it is not surprising that the cardinal points could serve an allegorical function in a diagrammatic or cartographic context. The close connection between geography and cosmology in medieval maps and diagrams has long been recognised.245See esp. Edson, Mapping Time and Space, 36–96; cf. Kedwards, Medieval Iceland, 63–100. The broader symbolic role of the cardinal points in these images has received significant attention, especially as it relates to number four symbolism, the representation of the cross and the acrostic reading of the Greek cardinal points as ‘Adam’.246Woodward, ‘Mappaemundi’, 335–7; Bronder, ‘Das Bild der Schöpfung’; Kochanek, Vorstellung vom Norden, 164–218; Wolever, ‘Post-Carolingian Geographies’, 163–7. This approach has emphasised how the cardinal points serve to ground the Christian order of the world, situating it inter alia within a relationship of microcosm and macrocosm or between creation and salvation, and establishing Christ’s governance over it. At the same time, the cosmological significance of the four cardinal winds provided similarly rich ground for cosmological analogies of four, eight and twelve.247Baumgärtner, ‘Winds and Continents’, 54–6; Thomas Raff, ‘Die Ikonographie der mittelalterlichen Windpersonifikationen’, Aachener Kunstblätter 48 (1978/9): 140–55. In particular, while rightly criticising the often overly narrow, spiritualising view of medieval intellectual life that previous, symbolist approaches to the cardinal points could fall into, Obrist highlights how the cardinal winds, and particularly their personifications, serve centrally to ground the coherence and stability of the cosmos in a very similar manner, both physical and spiritually.248Obrist, ‘Wind Diagrams’, 34, 70–84. Cf. Gautier Dalché, ‘Considérations intempestives’. All these factors provide important context for understanding why the cardinal points, especially around a wind diagram, could be seen as an appropriate context for allegorical commentary. There is, however, more that can be said about the cosmological function of the cardinal points around wind diagrams that will further help to explain why allegorisations would be added as glosses in this context.
Unlike the cardinal winds, the cardinal points are not typically an integral part of a wind diagram. Rather, they serve more often as a gloss external to the diagram as such, which can be included or not without fundamentally changing the diagram. To take the earliest Isidorian diagrams as an example, this most basic form of a wind diagram contains simply the twelve winds in Greek and Latin, set within the twelve spokes of a wheel.249Obrist, ‘Wind Diagrams’, 45, and fig. 4 (Munich, Bayerische Staatsbibliothek, Clm 16128, 35v). A range of further elements could then be added to this basic form to create new more complicated diagrams. For example, a number of early medieval wind diagrams include the words ‘kosmos’ and ‘mundus’ in the central circle of the diagram written perpendicular so that they intersect in a cross, a diagram associated especially with Vegetius manuscripts, but also Isidore’s De natura rerum.250Obrist, ‘Wind Diagrams’, 45–9, and, e.g., fig. 3 (Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 2077, 99r) and fig. 7 (Bern, Burgerbibliothek, Cod. A 92.20, 2r). Along the same lines, the inscription of a T-O diagram within a traditional Isidorian wind diagram came into fashion around the ninth century.251Obrist, ‘Wind Diagrams’, 49–53, and figs 9 and 10 (Bern, Burgerbibliothek, Cod. 212, 109r, Gautier Dalché, MM, Bern8; Berlin, Staatsbibliothek, MS Phill. 1830, 3v, Gautier Dalché, MM, Berl21). The cardinal points are not only more frequently found around these latter styles of wind diagrams that contain a geographical or cosmological ‘core’, but their function on these diagrams is likewise more clearly defined. In the Vegetian style diagrams (fig. 7), the cardinal points are not simply written around the outside of the circle, but they are embedded within the outer frame of the diagram so that the winds literally emerge from them.252Obrist, ‘Wind Diagrams’, 47, and figs 3 and 8, cf. 19 (Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 2077, 99r; Basel, Universitätsbibliothek, MS F III 15a, 22r; cf. Cologne, Erzbischöfliche Diözesan- und Dombibliothek, MS 83 II, 84r). Likewise for Isidorian wind roses with internal T-O diagrams (fig. 8), not only are the cardinal points often included, but they are sometimes connected directly to the diagram with boxes or other illustrations.253Obrist, ‘Wind Diagrams’, 52, and fig. 10 (Berlin, Staatsbibliothek, MS Phill. 1830, 3v); see also Karlsruhe, Badische Landesbibliothek, Cod. Aug. perg. 106, 52v (Gautier Dalché, MM, Kar2); Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 1260, 39r (Gautier Dalché, MM, VatR20). Thus, on these styles of diagram, the directions provide more than just contextual information, but serve to physically anchor both the winds and the cosmos.
Content - style instruction: <INSERT FIGURES 7 AND 8 NEAR HERE>
Despite individual scribes directly integrating the cardinal points into certain wind diagrams, no standard models emerge. The cardinal points remain glosses, added or elaborated upon by individual authors, but nevertheless external to the diagrams as such. This is especially evident in the case of two near identical tidal diagrams from Paris and Strasbourg, which have had an internal T-O diagram within the circle of winds inscribed in the middle.254Paris, Bibliothèque nationale de France, Latin 5239, 139r; Strasbourg, Bibliothéque nationale et universitaire, MS 326, 120v (Gautier Dalché, MM, Stra2). See Obrist, ‘Wind Diagrams’, 52. In contrast with the Paris manuscript, which is more typical of this style of diagram, the Strasbourg manuscript has added the four cardinal winds, with a short note explaining their order.255Cf. Obrist, ‘Wind Diagrams’, 52n89. More interesting still, the scribe has not only added the cardinal points to the diagram, but also adds on the east: ‘here the east rises’.256Strasbourg, Bibliothéque nationale et universitaire, MS 326, 120v: ‘hic oriens oritur’. Similarly in Paris, Bibliothèque nationale de France, Latin 13953, 72v (Gautier Dalché, MM, Par72). The paratextual function, therefore, of this inscription highlights how the cardinal points represent more of a gloss than an internal element of the diagram itself. They serve to orient the reader and offer both pedagogical and cosmological context to the diagram’s content, which allows for considerable flexibility when it comes to the content of these glosses.
Given this paratextual function, it may not surprise us that our allegorisations are not the first instance of such commentary-like glosses being added to cosmological diagrams. We find an earlier example, appearing once again in Echternach, on Willibrord’s horologium of the mid-eighth century.257Paris, Bibliothèque nationale de France, Latin 10837, 42r. On this illustration, with a transcription, see Barbara Obrist, ‘The Astronomical Sundial in Saint Willibrord’s Calendar and its Early Medieval Context’, Archives d’histoire doctrinale et littéraire du Moyen Age 67 (2000): 71–118. Unlike most later horologium diagrams, this diagram contains a range of paratextual elements outside the diagram itself, including various excerpts from Isidore’s Etymologies and three suns illustrated and labelled in different scripts.258Obrist, ‘Astronomical Sundial’, 76–8, notes that one is Latin and another Greek, but cannot identify the third. Importantly for us, one such paratext is the cardinal points. First the Greek cardinal points are labelled on the four sides of the diagram. Then there is a short description of the Latin cardinal points, each beginning with an etymology drawn from Isidore and then associated with one of four ages of man. While the east, south and west are relatively unremarkable, representing birth, youth and death respectively, the north takes on a more religiously charged interpretation: ‘The north signifies the Antichrist afflicting the world to old age’.259Obrist, ‘Astronomical Sundial’, 77: ‘[Aquilo] significat antichristum mundum stringentem ad senectutem’. The relationship here to Froumund’s application of Rabanus Maurus is certainly too loose to draw a causal connection, but taken alongside Willibrord’s horologium, we can see how the paratextual role of the cardinal points around a diagram or map provided a clear canvas for cosmological commentary, situating and orienting its contents both literally and spiritually.
Returning to where we started, there is a final important thread linking these cosmological diagrams back to our Sallust maps. The greater part of these texts are educational in focus. Indeed, the Echternach wind-diagram is already bound together with a copy of Sallust alongside other traditional school texts. Although neither the T-O diagram nor the Macrobian map found in that manuscript contain any further glosses related to the cardinal points, it is not hard to see how these materials bump up against one another in this sort of educational context.260Paris, Bibliothèque nationale de France, Latin 10195, 33v and 57r. And this is, of course, precisely the context of our original Sallust collection, where we have direct evidence for the continued use of at least one of these manuscripts in a school from the twelfth to fourteenth century.261Milan, Biblioteca Ambrosiana, G 21 Sup. See Black, ‘Rise and Fall of the Latin Classics’, 432. Thus the inclusion of these diagrams is not just offering a spiritualising, cosmological framework for contemplating the world, but it may also be serving to contextualise the study of Latin classics in the schoolroom around the study of the Bible.
Conclusion
As should be evident by now, the cardinal points were fundamental categories by which authors of the central Middle Ages conceptualised the world in which they lived. The significance and use of the cardinal points was in many ways tied to the rise of Christianity in late antiquity. Likewise, many geographical innovations during the Middle Ages grew out of the interplay of these new models of structuring the world and geographical texts with the lingering significance of ancient geographical models, like the three parts of the world. In tandem with the newly shifting boundaries, especially for Africa and Asia, the cardinal points carved out an increasingly prominent and autonomous position within the conceptualisation of geographical space. They served not only as a tool to clarify previous ambiguities in the geographical tradition, but also to highlight ruptures that were developing in medieval world geography. Hence both the north-west and the south-east corners of the world become contested regions, as they fell between these different geographical groupings. Similarly, the literary formulae of late antiquity ran up against new ways of reading geographical space in the Middle Ages. While the interplay between the different models continued throughout the Middle Ages (and continue for us today), it is clear that by the twelfth century, the cardinal points had become the basic vocabulary within this discourse by which to conceptualise, structure and contextualise the world and according to which other systems were themselves understood.
 
1     Martin W. Lewis and Kären E. Wigen, The Myth of Continents: A Critique of Metageography (Berkeley: University of California Press, 1997), 23–4; for an accurate and nuanced treatment of this issue, see Mauntel, Erdteile. »
2     On medieval geographical innovations, see already Raymond Beazley, The Dawn of Modern Geography, 3 vols (London: John Murray, 1897–1906), 2:8–13. »
3     Mauntel, Erdteile, 77–8, 282–5, 580, passim. »
4     Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 38, trans. Charles C. Mierow (New York: Columbia University Press, 2002), 123: ‘Quarum provintias, situs, regiones qui velit cognoscere, legat Orosium’ (If anyone wishes to know the provinces, the topography and the divisions of these parts of the world, let him read Orosius). Cf. Hon., Im. mun. prol. Christ., ed. Flint, 49: ‘Nichil autem in eo pono nisi quod majorum commendat traditio’ (I have set down nothing in this book except what the tradition of our ancestors has commended). On Signot’s La division du monde, see Mauntel, Erdteile, 580. »
5     Felicitas Schmieder, ‘Gogs und Magogs “natürliche Milde”? Die Mongolen als Endzeitvölker im Wandel von Wissen und Wünschen’, in Peoples of the Apocalypse: Eschatological Beliefs and Political Scenarios, ed. Wolfram Brandes, Felicitas Schmieder and Rebekka Voß (Berlin: De Gruyter, 2016), 112, and compare Schmieder’s treatment of Gog and Magog with Remi of Auxerre’s letter about the Hungarians in R. B. C. Huygens, ‘Un témoin de la crainte de l’an 1000: la lettre sur les Hongrois’, Latomus 15, no. 2 (1956): 225–39, esp. ll. 96–101, 161–201; cf. Claude Nicolet and Patrick Gautier Dalché, ‘Les “quatre sages” de Jules César et la “mesure du monde” selon Julius Honorius: réalité antique et tradition médiévale’, Journal des savants 4, no. 1 (1986): 197–8n53. See also Wolever, ‘Post-Carolingian Geographies’, 153–63. »
6     Natalia Lozovsky, ‘The Earth Is Our Book’: Geographical Knowledge in the Latin West ca. 400–1000 (Ann Arbor: University of Michigan Press, 2000), ch. 1; Deluz, ‘Une image du monde’, 44. »
7     Deluz, ‘Une image’, 41. On the expansion of education in the period, see Ferroulo, Origins of the University. On the expansion of monasteries, see Giles Constable, The Reformation of the Twelfth Century (Cambridge: Cambridge University Press, 1996), 88–93. »
8     Hon., Im. mun. prol. Christ., ed. Flint, 48–9; HoSV, Desc. prol., 12, ed. Gautier Dalché, 133; Patrick Gautier Dalché, ‘Nouvelles lumières sur la Descriptio mappe mundi de Hugues de Saint-Victor’, in Géographie et culture: La représentation de l’espace du VIe Au XIIe siècle (Farnham: Ashgate, 1997), 1–27; cf. William M. Green, ‘Hugo of St Victor: De Tribus Maximis Circumstantiis Gestorum’, Speculum 18, no. 4 (1943): 484–93. On Guido of Pisa, see Campopiano, ‘Construction of the Text’, 82. »
9     P. D. A. Harvey, ‘The Sawley Map and Other World Maps in Twelfth-Century England’, Imago Mundi 49, no. 1 (1997): 37–8; Dan Terkla, ‘Hugh of St Victor (1096–1141) and Anglo-French Cartography’, Imago Mundi 65, no. 2 (2013): 161–79; Dan Terkla, ‘Introduction: Where to fix Cadiz?’, and Nathalie Bouloux, ‘The Munich Map (c. 1130): Description, Meanings and Uses’, in A Critical Companion to English Mappae Mundi of the Twelfth and Thirteenth Centuries, ed. Dan Terkla and Nick Millea (Woodbridge: Boydell, 2019), 10–11, 93. »
10     Lozovsky, Earth Is Our Book, 6–10; Patrick Gautier Dalché, La géographie de Ptolémée en Occident (IVe–XVIe siècle) (Turnhout: Brepols, 2009). »
11     For example, see the confusion over Thule, Balcia and Scandinavia (Fraesdorff, Norden, 46–50). »
12     James Romm, ‘Continents, Climates, and Cultures: Greek Theories of Global Structure’, in Geography and Ethnography: Perceptions of the World in Pre-Modern Societies, ed. Kurt A. Raaflaub and Richard J. A. Talbert (Chichester: Wiley-Blackwell, 2010), 215–20. The Tanais and Phasis correspond to the modern rivers Don and Rioni. »
13     Herodotus, Histories 4.35, 42, trans. Robin Waterfield (Oxford: Oxford University Press, 1998), 247–8; see Rosalind Thomas, Herodotus in Context (Cambridge: Cambridge University Press, 2002), ch. 3, and Reinhold Bichler, ‘Der Antagonismus von Asien und Europa – eine historiographische Konzeption aus Kleinasien?’, in Der Beitrag Kleinasiens zur Kultur- und Geistesgeschichte der griechisch-römischen Antike, ed. Josef Fischer (Vienna: Austrian Academy of Sciences, 2014), 10–11. »
14     Germaine Aujac, ‘The Foundations of Theoretical Cartography in Archaic and Classical Greece’, in The History of Cartography, vol. 1, ed. J. B. Harley and David Woodward (Chicago: University of Chicago Press, 1987), 135–7. »
15     Romm, ‘Continents, Climates, and Cultures’, 218–20; Isaac, Invention of Racism, 56–60. For a more extensive discussion of the ways in which a north-south dichotomy takes precedence over east-west, see James Romm, ‘Herodotus and Mythic Geography: The Case of the Hyperboreans’, Transactions of the American Philological Association 119 (1989): 97–113. »
16     Isaac, Invention of Racism, ch. 4; cf. Bichler, ‘Der Antagonismus’, 12–14. »
17     Isaac, Invention of Racism, 60n18; Thomas, Herodotus, 86. »
18     Airs, Waters, Places 12, trans. Paul Potter, Hippocrates, vol. 1, Loeb Classical Library 147 (Cambridge, MA: Harvard University Press, 2022), 109; Isaac, Invention of Racism, 61. »
19     Romm, ‘Continents, Climates, and Cultures’, 220–3, see 232n21 for discussion on whether Libya is treated as a continent in its own right. »
20     Airs, Waters, Places 12–13, trans. Potter, 109–15. »
21     Romm, ‘Continents, Climates, and Cultures’, 221–3. »
22     Isaac, Invention of Racism, 283–5; Romm, ‘Continents, Climates, and Cultures’, 223–6; cf. Bichler, ‘Der Antagonismus’, 14–16. »
23     Romm, ‘Continents, Climates, and Cultures’, 225–6. »
24     Romm, ‘Continents, Climates, and Cultures’, 223–4, 226. »
25     Plato, Republic 4.435e–436a, trans. G. M. A. Grube, in Plato: Complete Works, ed. John M. Cooper (Indianapolis: Hackett, 1997), 1067; Aristotle, Politics 7.7.1327b, trans. Jonathan Barnes, The Complete Works of Aristotle, 2 vols (Princeton: Princeton University Press, 1984), 2:2107; cf. Isaac, Invention of Racism, 73. »
26     Isaac, Invention of Racism, 283–98; Romm, ‘Continents, Climates, and Cultures’, 224–6. »
27     Romm, ‘Continents, Climates, and Cultures’, 226–7; Strabo, Geography 11.7.4, cf. 11.5.5, trans. Horace Leonard Jones, 8 vols, Loeb Classical Library 49, 50, 182, 196, 211, 223, 241, 267 (Cambridge, MA: Harvard University Press, 1917–32), 5:255, cf. 5:239. »
28     E.g., Isaac, Invention of Racism, 83–5; Guy Halsall, Barbarian Migrations and the Roman West 376–568 (Cambridge: Cambridge University Press, 2007), 46–7. »
29     E.g., Varro, Res rusticae 1.2.3–4, ed. Georg Goetz (Leipzig: Teubner, 1929), 10: ‘cum orbis terrae divisus sit in duas partes ab Eratosthene maxume secundum naturam, ad meridiem versus et ad septemtriones …’ (Since the earth was most naturally divided into two parts by Eratosthenes, towards the south and toward the north …); Strabo, Geography 2.1.1, trans. Jones, 1:223. »
30     Polybius, Histories 3.37.2–8, trans. W. R. Paton, vol. 2, Loeb Classical Libary 137 (Cambridge, MA: Harvard University Press, 2010), 95–7. »
31     Varro, De lingua latina 5.5.31, ed. Georg Goetz and Fritz Schöll (Leipzig: Teubner, 1910), 11: ‘Ut omnis natura in caelum et terram divisa est, sic caeli regionibus terra in Asiam et Europam. Asia enim iacet ad meridiem et austrum, Europa ad septemtriones et aquilonem’ (As the whole of nature is divided into earth and sky, so the earth is divided by the regions of the sky into Asia and Europe. For Asia lies to the south and south-wind while Europe lies to the north and north-wind). Cf. Varro, Res rusticae 1.2.3–4, ed. Goetz, 10. »
32     Pliny, Naturalis historia 3.1.5, ed. and trans. Rackham, 2:4–7: ‘Primum ergo de Europa, altrice victoris omnium gentium populi longeque terrarum pulcherrima, quam plerique merito non tertiam portionem fecere, verum aequam, in duas partes ab amne Tanai ad Gaditanum fretum universo orbe diviso’ (To begin then with Europe, nurse of the race that conquered all the nations, and by far the loveliest portion of the earth, which most authorities, not without reason, have reckoned to be not a third part but a half of the world, dividing the whole circle into two portions by a line drawn from the river Don to the Straits of Gibraltar). Cf. Polybius, Histories 3.37.4, trans. Paton, 95, and Strabo, Geography 2.4.5, 11.7.4, trans. Jones, 1:411–13, 5:255. »
33     Claudia Wick, M. Annaeus Lucanus Bellum civile liber IX Kommentar (Munich: K. G. Saur, 2004), 159–60; Benet Salway, ‘Putting the World in Order: Mapping in Roman Texts’, in Ancient Perspectives: Maps and Their Place in Mesopotamia, Egypt, Greece, and Rome, ed. R. J. A. Talbert (Chicago: University of Chicago Press, 2012), 213–14; cf. Mauntel, Erdteile, 102–3, who remains sceptical about Asia’s geographical priority in the Roman period. »
34     Sallust, Bellum Iugurthinum 17.3, ed. Alfons Kurfess (Leipzig: Teubner, 1957), 68: ‘In divisione orbis terrae plerique in parte tertia Africam posuere, pauci tantummodo Asiam et Europam esse, sed Africam in Europa’ (In the division of the world, the majority consider Africa as a third part, while a few maintain there is only Asia and Europe, but set Africa within Europe). »
35     Pomponius Mela, De chorographia 1.9, ed. Alain Silberman (Paris: Les Belles Lettres, 1988), 3–4; Strabo, Geography 17.3.1, trans. Jones, 8:155. In the latter, Strabo seems to imply a potential combination of Europe and Africa when he suggests that even combined with Europe, Africa would not equal Asia’s extent. Pace Merrills’s restriction of such an association of Africa and Europe to Sallust (Andrew Merrills, History and Geography in Late Antiquity (Cambridge: Cambridge University Press, 2005), 76–7). »
36     Lucan, 9.411–3, ed. Shackleton Bailey, 240, trans. Susan H. Braund (Oxford: Oxford University Press, 1992), 118 [hereafter Lucan]: ‘Tertia pars rerum Libye, … / at, si ventos caelumque sequaris, / pars erit Europae’. »
37     Philip Hardie, Vergil’s Aeneid: Cosmos and Imperium (Oxford: Clarendon Press, 1986), 311–13. »
38     Augustine, De civitate Dei 16.17, ed. B. Dombart and A. Kalb, CCSL 47–8 (Turnhout: Brepols, 1955), 521; Augustine, Questionum in heptateuchum 6.15, ed. J. Fraipont, CCSL 33 (Turnhout: Brepols, 1958), 321; Orosius, 1.2.1, ed. Zangemeister, 5; Isi., Etym. 14.2.2. Similarly among pagan authors of this period as, e.g., Servius, In Vergili Aeneidos 1.385, ed. Thilo, 1:130. »
39     Patrick Gautier Dalché, ‘L’enseignement de la géographie dans l’antiquité tardive’, Klio 96, no. 1 (2014): 150–1; H. I. Marrou, A History of Education in Antiquity, trans. George Lamb (New York: Mentor, 1956), 373–5; Alan Cameron, Greek Mythology in the Roman World (Oxford: Oxford University Press, 2004), 3–4. »
40     Wick, Lucanus, 160. »
41     There are exceptions, notably Varro, De lingua latina 5.5.31, ed. Goetz and Schöll, 11. See more broadly Nicolet, ‘Les “quartre sages”’, 180–1n63. »
42     Pomponius Mela, De chorographia 1.8, ed. Silberman, 3, trans. F. E. Romer, Pomponius Mela’s Description of the World (Ann Arbor: University of Michigan Press, 1998), 36: ‘hoc mari et duobus inclutis amnibus, Tanai atque Nilo, in tres partes universa dividitur’. Cf. Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Sallust, Bellum Iugurthinum 17.3, ed. Kurfess, 68; Strabo, Geography 1.2.1, 2.5.18, trans. Jones, 1:49–51, 467–9. »
43     Augustine, De civitate Dei 16.17, CCSL 48, 521, trans. R. W. Dyson (Cambridge: Cambridge University Press, 1998), 725: ‘ista, quae Asia nuncupatur, a meridie per orientem usque ad septentrionem peruenit; Europa uero a septentrione usque ad occidentem, atque inde Africa ab occidente usque ad meridiem’. »
44     E.g., Gn 13 and Lk 13:29. Compare the use of water boundaries in the not overtly Christian Martianus Capella, De nuptiis 6.622–3, ed. Willis, 217. »
45     For a survey of medieval formulae, see Goetz, Gott und die Welt, vol. 1, bk. 2, 136–40; cf. Mauntel, Erdteile, 156, and Christoph Mauntel, ‘The T-O Diagram and its Religious Connotations: A Circumstantial Case’, in Geography and Religious Knowledge in the Middle Ages, ed. Christoph Mauntel (Berlin: De Gruyter, 2021), 79–80. »
46     Pliny, Naturalis historia 2.46.119, ed. Rackham, 1:260–1: ‘per totidem mundi partes’; ‘quattuor caeli partes’. »
47     Pomponius Mela, De chorographia 1.3–4, ed. Silberman, 2; cf. Isi., Etym. 3.42. »
48     Alfred Hiatt, Terra Incognita: Mapping the Antipodes Before 1600 (Chicago: University of Chicago Press, 2008), 16–17; Wesley M. Stevens, ‘The Figure of the Earth in Isidore’s “De natura rerum”’, Isis 71, no. 2 (1980): 269–70; Germaine Aujac, ‘L’image du globe terrestre dans la Grèce ancienne’, Revue d’histoire des sciences 27, no. 3 (1974): 205–10. See Martianus Capella, De nuptiis 6.604–7, ed. Willis, 211–12. »
49     Agatharchides, De mari Erythraeo 64, trans. Karl Müller, Geographi Graeci minores, vol. 1 (Paris: Ambroise Firmin Didot, 1855), 156; see Nicolet, ‘Les “quartre sages”’, 180–1n63–5. »
50     Nicolet, ‘Les “quatre sages”’, 162–3. On the complex manuscript tradition, the question of two ‘recensions’ and the relationship of Julius Honorius to Julius Orator, see Gautier Dalché, ‘Les “quatre sages”’, 184–92. »
51     Gautier Dalché, ‘Les “quatre sages”’, 187–8; Cassiodorus, Institutiones 1.25, ed. R. Mynors (Oxford: Clarendon Press, 1961), 66. »
52     These later copies lack any reference to Julius Honorius or Julius Orator and are attributed instead to Julius Caesar in earlier manuscripts: ‘Cosmographia Iulii Caesaris’ or ‘Cronica Iulii Caesaris’ (Alexander Riese, Geographi latini minores (Heilbronn: Henninger, 1878), 21). Besides Cassiodorus, the original title: ‘Cosmografia’ appears in Paris, Bibliothèque nationale de France, Latin 2769, 23v. On its relationship to Paris, Bibliothèque nationale de France, Latin 4808, see Gautier Dalché, ‘Les “quatre sages”’, 190, and E. A. Lowe, Codices Latini Antiquiores, 12 vols (Oxford: Clarendon Press, 1934–71), 5:10, no. 550. »
53     This is not to be confused with the Cosmographia of Aethicus Ister, edited by Michael W. Herren, The Cosmography of Aethicus Ister: Edition, Translation, and Commentary (Turnhout: Brepols, 2011), and Otto Prinz, Die Kosmographie des Aethicus, MGH QQ zur Geistesgesch. 14 (Munich: MGH, 1993). On the relationship of these texts, see Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 161n10 and 198–9. »
54     Gautier Dalché, ‘Les “quatre sages”’, 192–8. »
55     Saint-Omer, Bibliothèque d’Agglomération, MS 717; Albert Derolez, The Autograph Manuscript of the Liber Floridus: A Key to the Encyclopedia of Lamber of Saint-Omer (Turnhout: Brepols, 1998), 193. »
56     A continued awareness of these conflicting models may be reflected in an inscription shared across three eleventh-century computus manuscripts: Dijon, Bibliothèque municipale, MS 448, 74v: ‘mundus quatuor angulos habet et partes tres diuisas’ (The world has four corners and three separate parts). Also Trier, Stadtbibliothek, Hs. 4° 1084/115, 99r and Boulogne-sur-Mer, Bibliothèque municipale, MS 188, 30r. For descriptions and reproductions, see Patrick Gautier Dalché, Mappae mundi (VIIIe–XIIe siècle), 2 vols (Turnhout: Brepols, 2024), Dij4, Boul2, Tri1 [hereafter Gautier Dalché, MM] (all citations follow Gautier Dalché’s alphanumeric code for each individual map). On these diagrams, see Obrist, ‘Wind Diagrams’, 58, and Bruce Eastwood, Ordering the Heavens: Roman Astronomy and Cosmology in the Carolingian Renaissance (Leiden: Brill, 2007), 416. »
57     Pseudo-Aethicus, Cosmographia 2.1, ed. Alexander Riese, Geographi latini minores (Heilbronn: Henninger, 1878), 90 [hereafter Ps.-Aeth.]: ‘Hanc quadripertitam totius terrae continentiam hi qui dimensi sunt longe citroque maiores nostri tripertitam reputari definierunt, investigantes universum orbem oceani maris limbo circumdatum, easque tres partes Asiam Europam et Africam reputarunt. Quamvis non defuerunt, qui duas partes sicut diximus perhiberent, Asiam et Europam; Africam vero in Europa adiciendam adfinierunt, quia et spatio latitudinis eget et caeli male subiacet climati, laborans aeribus suis, venenis †fucisque repleta inmanium et incognitarum humano generi innumerabilium bestiarum. Sed ad propositum remeantes absque illorum praeiudicio, qui hoc omne totum duas partes esse voluerunt, tripertitam divisionem dicere incipiamus’. This text is based closely on Orosius, 1.2.1, ed. Zangemeister, 5, trans. Fear, 36: ‘Maiores nostri orbem totius terrae, oceani limbo circumsaeptum, triquadrum statuere eiusque tres partes Asiam Europam et Africam uocauerunt, quamuis aliqui duas hoc est Asiam ac deinde Africam in Europam accipiendam putarint’ (Our ancestors divided the whole world, surrounded as it is by the belt of the Ocean, into three rectangular blocks, and called these three parts Asia, Europe, and Africa, although there are some who believe that there are two parts: namely Asia and Europe, including Africa in the latter part). »
58     Tres partes: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Lucan, 9.411, ed. Shackleton Bailey, 240; Isi., Etym. 14.2.1–2; Orosius, 1.2.1, ed. Zangemeister, 5; Ps.-Aeth., 2.1, ed. Riese, 90; HoSV, Desc. 7, ed. Gautier Dalché, 138; Hon., Im. mun. 1.7, ed. Flint, 52; Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 37; cf. Tres Climata: OV, 11.12, ed. and trans. Chibnall, 6:70–1; Quattuor climata: Isi., Etym. 13.1.3; Hon. Im. mun. 1.91, ed. Flint, 81; Bede, De natura rerum 10, CCSL 123A, 200–1; cf. Quattuor Partes in Pliny, Naturalis historia 2.46.119, ed. Rackham, 1:260–1; Quattuor Cardines in Orosius, 2.1.5, ed. Zangemeister, 35. See also Mauntel, Erdteile, 42–4. »
59     Patrick Gautier Dalché, ‘Maps in Words: The Descriptive Logic of Medieval Geography, from the Eighth to the Twelfth Century’, in The Hereford World Map: Medieval World Maps and their Context, ed. P. D. A. Harvey (London: British Library, 2006), 223–42, and Patrick Gautier Dalché, ‘Comment et pourquoi décrire une mappemonde au Moyen Âge?’, in Figures de l’autorité médiévale: Mélanges offerts à Michel Zimmermann, ed. Pierre Chastang, Patrick Henriet and Claire Soussen (Paris: Publications de la Sorbonne, 2016), 69–88. »
60     Gautier Dalché, ‘L’enseignement de la géographie’, 158, 164; Cassiodorus, Institutiones 1.25, ed. Mynors, 66. »
61     Nathalie Bouloux, ‘L’espace habité’, in Gautier Dalché, ed., La terre, 259–60. Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 157–218, esp. 203–5; Guido of Pisa, Liber Guidonis, ed. Michele Campopiano (Florence: SISMEL, 2008), 21–2; Lamb., Lib. flor., 241r (Gautier Dalché, MM, Gent13). »
62     Jul. Hon., Cosm. 1, ed. Riese, 21: ‘Iulio Caesare et Marco Antonio consulibus omnis orbis peragratus est per sapientissimos et electos viros quattuor: Nicodemo orientis, Didymo occidentis, Theudoto Septemtrionalis, Polyclito meridiani’. On the inclusion of this story in later copies of Julius Honorius, see Gautier Dalché, ‘Les “quatre sages”’, 185–6. »
63     Ps.-Aeth., 1.1*–2*, ed. Riese, 72: ‘Itaque lulius Caesar bissextilis rationis inventor divinis humanisque rebus singulariter instructus cum consulatus sui fasces regeret, ex senatus consulto censuit omnem orbem iam Romani nominis admetiri per prudentissimos viros et omni philosophiae munere decoratos’ (And so Julius Caesar, inventor of the leap year and singularly learned in things human and divine, when he held his consulship, resolved on degree of the senate that the whole world now of Roman name be measured by men most skill and distinguished by every field of philosophy). »
64     Hartmut Kugler, Die Ebstorfer Weltkarte, 2 vols (Berlin: Akademie, 2007), 1:42: ‘Quam [i.e. mappa mundi] Julius Cesar missis legatis per totius orbis amplitudinem primus instituit’. »
65     Scott D. Westrem, The Hereford Map: A Transcription and Translation of the Legends with Commentary (Turnhout: Brepols, 2001), 9, trans. Westrem: ‘Lucas in evvangelio: Exiit edictum ab Augusto Cesare, ut describeretur huniversus orbis’. »
66     Westrem, Hereford Map, 9–11. »
67     Westrem, Hereford Map, 3, my trans.: ‘A Julio Cesare orbis terrarum metiri cepit: A Nicodoxo omnis Oriens dimensus est. A Theodoco Setpemtrion et Occidens dimensus est. A Policlito Meridiana pars dimensus est’ (emphasis added). Cf. Ps.-Aeth., 1.1*–2*, ed. Riese, 72. »
68     Derolez, The Autograph Manuscript, 193. »
69     Lamb., Lib. flor., 19r (Gautier Dalché, MM, Gent2); cf. Derolez, The Autograph Manuscript, 49. »
70     Ps.-Aeth., 1.48, ed. Riese, 90. »
71     Lamb., Lib. flor., 19r: ‘Asia Septentrionalis habet gentes’; Ps.-Aeth., 1.38, ed. Riese, 87–8. »
72     Lamb., Lib. flor., 19r: ‘Oriens habet’. »
73     Lamb., Lib. flor., 241r: ‘Europe, one fourth part of the world’. »
74     Lamb., Lib. flor., 241r: ‘nominatur pars tercia, sed uere est quarta’ ([Europe] is called the third part, but it is really one fourth part). Note that Latin ordinals like quartus can mean both the fourth (in a series) and one fourth (of a whole). »
75     Ps.-Aeth., 1.2*, ed. Riese, 72. »
76     Lamb., Lib. flor., 138v: ‘Deinde regna orientis et meridiani ac septentrionis occidentisque et prouintias insulasque per tres uiros prudentissimos uidelicet Nicodoxum et Pollyclitonem atque Theodotum metiri precepit’ (Then [Julius Caesar] ordered that the kingdoms of the east, south, and north and west as well as the provinces and islands be measured by three most learned men, namely Nicodemus, Polyclitus and Theodotus). Gautier Dalché, ‘Les “quatre sages”’, 203–5, 209. »
77     Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–70; cf. Fraesdorff, Norden, ch. 3. »
78     E.g., Orosius, 1.2.2, ed. Zangemeister, 5; trans. Fear, 36: ‘Asia … per totam transuersi plagam orientis extenditur’ (Asia … extends across the entire east). Cf. HoSV, Desc. 7, ed. Gautier Dalché, 138, and Nicolet, ‘Les “quatre sages”’, 181. »
79     Romm, Edges of the Earth, 45–60. »
80     Orosius, 1.2.9, 15, 27–34, 87–92, ed. Zangemeister, 5–8, 13–14. »
81     Isi., Etym. 14.3.27, 14.5.3. »
82     Jul. Hon., Cosm. 43–4, ed. Riese, 47; Ps.-Aeth., 1.39–46, ed. Riese, 88: ‘Aegyptum, Aethiopiam, Africam’. »
83     Dicuil, Liber de mensura orbis terrae prol.6, 4, ed. Tierney, 44–5, 54–7; cf. Situs orbis 4.13, ed. Patrick Gautier Dalché, ‘Situs orbis terre vel regionum: Un traité de géographie inédit du haut Moyen Âge (Paris, B. N. latin 4841)’, Revue d’histoire des textes 12–13 (1982–3): 167, and Ravennatis anonymi cosmographia 2.20–1, ed. Joseph Schnetz, Intineraria Romana, vol. 2 (Leipzig: Teubner, 1990), 32–3. Confusion around India’s relationship to Ethiopia also continued into the later Middle Ages. E.g., Gervase of Tilbury, Otia imperialia 2.3, ed. and trans. S. E. Banks and J. W. Binns (Oxford: Clarendon Press, 2002), 183; cf. Bernard Hamilton, ‘Continental Drift: Prester John’s Progress through the Indies’, in Prester John, the Mongols and the Ten Lost Tribes, ed. Charles F. Backingham and Bernard Hamilton (Aldershot: Variorum, 1996), 238–40, and Rouxpetel, L’Occident au miroir, 115–37. »
84     Flint in Hon., Im. mun. 7–8; more generally on Honorius, see the collected essays in Valerie I. J. Flint, Ideas in the Medieval West: Texts and Their Contexts (London: Ashgate, 1988). For surveys on Honorius’s life see Karl Kinsella, ‘Edifice and Education: Thought in Twelfth-Century Europe’, 2 vols (PhD diss., University of Oxford, 2016), 1:214–18, and Fabian Schwarzbauer, Geschichtszeit: Über Zeitvorstellungen in den Universalchroniken Frutolfs von Michelsberg, Honorius’ Augustodunensis und Ottos von Freising (Berlin: Akademie, 2005), 40–2. »
85     Hon., Im. mun. 1.17, ed. Flint, 57: ‘Hee superius dicte regiones ab oriente incipientes, recta linea ad Mediterraneum mare extenduntur, quibus versus Austrum Egyptus connectitur’. »
86     Cf. Isi., Etym. 14.3.27. »
87     Ps.-Aeth., 1.5, 43, ed. Riese, 73, 88. Flint cites no influence of the Cosmographia on the Imago Mundi (Flint in Hon., Im. mun. 46–7, passim). »
88     Gautier Dalché in HoSV, Desc. 171n43, and ‘Maps in Words’, 230–1. This is clearest on the Munich Map (Munich, Bayerische Staatsbibliothek, CLM 10058, 154v), but see also the Sawley Map (Cambridge, Corpus Christi College, MS 66, 1v). Gautier Dalché, MM, Mün27, Camb2. On these maps, see the chapters by Nathalie Bouloux, ‘The Munich Map’, and Alfred Hiatt, ‘The Sawley Map (c. 1190)’, in Terkla and Millea, eds, Critical Companion to English Mappae Mundi. »
89     Hon., Im. mun. 1.30, ed. Flint, 63: ‘Hec in oriente Indi fluminis surgit, et per meridiem vergens, in occidentem tendit’. The earliest family of manuscripts, C, has ‘Nili’ in place of ‘Indi’. »
90     This can be seen, e.g., on both the Munich and Sawley Maps (Munich, Bayerische Staatsbibliothek, CLM 10058, 154v, and Cambridge, Corpus Christi College, MS 66, 1v). »
91     Isi., Etym. 14.2.2, my trans.: ‘Asia a meridie per orientem usque ad septentrionem pervenit’. »
92     Isi., Etym. 14.2.3. »
93     Hon., Im. mun. 1.32, ed. Flint, 64: ‘Versus meridiem vero est Ethiopia, ab Etham dicta, una in oriente … altera in occidente … Ultra Ethiopiam sunt maxima loca deserta … Deinde est maximus oceanus’. »
94     Gautier Dalché in HoSV, Desc. 55–8. On Hugh more generally, see Paul Rorem, Hugh of Saint Victor (Oxford: Oxford University Press, 2009), ch. 1. »
95     HoSV, Desc. 18, ed. Gautier Dalché, 151: ‘sed sciendum quod scriptores aliquando Libiam pro Affrica, et Affricam pro Libia ponunt, sed et Ethiopiam, Egyptum, Libiam, Affricam pro meridiana plaga indifferenter ponunt’. »
96     Cf. Gautier Dalché in HoSV, Desc. 51n34, 171n43. »
97     Paris, Bibliothèque nationale de France, Latin 8674, 71v, and Paris, Bibliothèque nationale de France, NAL 340, 77v: ‘Lybicum i.e. Africum i.e. australem’. On Martianus Capella glosses, see Lozovsky, Earth Is Our Book, 113–38, esp. 116, and Cora E. Lutz, ‘Martianus Capella’, in Catologus translationum et commentorum, vol. 2, ed. Paul Oskar Kristeller (Washington, DC: Catholic University of America Press, 1971), 372–6. »
98     Peter the Venerable, Liber contra sectem Saracenorum prol., D 179vs, ed. James Kristzeck, Peter the Venerable and Islam (Princeton: Princeton University Press, 1964), 226 (emphasis added): ‘toto fere ut dictum est armis oriente subacto, Egyptum, Lybiam, Affricamque uniuersam prophanae religioni subiecit, et sic duabus mundi partibus occupatis’ (see p. 000 below for a translation); Guibert, Dei gesta 1.4, CCCM 127A, 98–9: ‘Huius nefariae institutionis obscuritas christianum tunc nomen obtexit et adhuc per Orientis pene universi, Affricae, Egypti, Ethiopiae, Libiae et iuxta nos Hispaniae remotissimos sinus oblitterat’ (The shadow of this wicked sect covered the Christian name and has now consigned it to oblivion through the furthest recesses of almost the whole east, of Africa, Egypt, Ethiopia, Libya and of Spain, next to us). »
99     HoSV, Desc. 9–12, cf. capitula, ed. Gautier Dalché, 140–5, cf. 133. N.B. ‘IX. De regionibus et gentibus Asye ad austrum’ is listed among the capitula, however in the text we find ‘IX. De prouinciis et ciuitatibus Asie ad ortum solis’, covering the geography of India, which includes Ethiopian India ending at the Red Sea in the south (on the regions and peoples of Asia to the south; on the provinces and cities of Asia to the east). »
100     HoSV, Desc. 9–10, 14, ed. Gautier Dalché, 140–1, 146. »
101     On the Chronica, see Mortensen, ‘Hugh of St. Victor on Secular History’, 3–30; Green, ‘De Tribus Maximis’, 484–93. For a complete transcription of the Chronica, see Hugonis de Sancto Victore operum Editio auspiciis Gilduini abbatis procurata et IV voluminibus digessa, ed. Rainer Berndt and José Luis Narvaja, vol. 1 (Münster: Aschendorff, 2017). As Berndt and Narvaja reproduce a single manuscript: Paris, Bibliothèque nationale de France, Latin 15009, all my citations follow that manuscript. »
102     Paris, Bibliothèque nationale de France, Latin 15009, 16r: ‘In latere australi asie porrigitur mare rubrum ab occeano orientali occidentem uersus usque ad confinia arabie et palestine. / Inter mare rubrum et occeanum australem ad orientem supra ethiopia est superior que ab occidente montibus ethiopicis per transuerssum terminatur. / Sub ethiopia orientali ad occidentem hee sunt regiones porrecte / Egiptus superior / Deserta egypta / Egyptus inferior / Lybia superior / In affrica sunt regiones iste / Ethiopes nadaberes [etc.]’. »
103     Cf. Leipzig, Universitätsbibliothek, MS 350, 107v, and Munich, Bayerische Staatsbibliothek, Clm 2, 101v. »
104     Guido of Pisa, Geographica 122, ed. Schnetz, 140: ‘Africa habet fines ab oriente praelatam Arabiam et regionem maris Rubri et ex parte regionem Hebraeorum’ (Africa has as boundaries on the east the aforementioned Arabia and the region of the Red Sea as well as in part the region of the Hebrews). Cf. Orosius, 1.2.8, ed. Zangemeister, 5; Isi., Etym. 14.5.3. On Guido, see Campopiano, ‘Construction of the Text’, 67–73; Campopiano in Guido of Pisa, Liber Guidonis, LI–LXII. »
105     Guido of Pisa, Geographica 121, ed. Schnetz, 139; cf. Orosius, 1.2.27–34, ed. Zangemeister, 7–8; Isi., Etym. 14.3.27. »
106     Gautier Dalché in HoSV, Desc. 171n43. »
107     Patrick Gautier Dalché, ‘L’héritage antique de la cartographie médiévale: Les problèmes et les acquis’, in Cartography in Antiquity and the Middle Ages: Fresh Perspectives, New Methods, ed. Richard Talbert (Leiden: Brill, 2008), 36–53, at 51: ‘“mise en cart” des textes geographiques’; Patrick Gautier Dalché, ‘Eucher de Lyon, Iona, Bobbio: Le destin d’une mappa mundi de l’antiquité tardive’, Viator 41 (2010): 1–22. »
108     See p. 000 above. »
109     Albi, Bibliothèque municipale, MS 29, 57v; ed. F. Glorie, Itineraria et alia geographia, CCSL 175 (Turnhout: Brepols, 1965), 467–71. Gautier Dalché, MM, Alb1. On the Albi map, see Emmanuelle Vagnon and Sandrine Victor, eds, La Mappa mundi d’Albi: Culture géographique et représentation du monde au haut Moyen Âge (Paris: Éditions de la Sorbonne, 2022). »
110     On the form, see Anca Dan, ‘La mappemonde d’Albi – Un pinax chôrographikos’, Revue du Comité français de cartographie 234 (2017): 27; Gautier Dalché, ‘L’enseignement de la géographie’, 163; and Priscian, Periegesis 10–13 (ed. Paul van de Woestijne, La Périégèse de Priscien: Édition critique (Bruges: De Tempel, 1953), 37). »
111     Orosius, 1.2.8, 34, ed. Zangemeister, 5, 8. On the influence of Orosius and Pseudo-Aethicus, see Konrad Miller, Mappaemundi: Die ältesten Weltkarten, vol. 3, Die kleineren Weltkarten (Stuttgart: Jos. Roth, 1895), 58–9, but see also the more nuanced discussion in Jean-Baptiste Amat, ‘La Mappa mundi d’Albi au regard de la Mappa mundi du Vatican comme illustration de la nécessaire comparaison des mappae mundi du haut Moyen Âge’, in Vagnon and Victor, eds, La Mappa mundi d’Albi, 180–1, 189–94. »
112     Dan, ‘La mappemonde d’Albi’, 23–7. »
113     For the stemma, see John Williams, ‘Isidore, Orosius and the Beatus Map’, Imago Mundi 49, no. 1 (1997): 10. Cf. Valladolid, Biblioteca Universitaria, MS 433, 36v–37r; Seu d’Urgell, Museo Diocesano, Num. Inv. 501, VIv–VIIr. »
114     New York, Morgan Library, MS M.644, 33v–34r. Reproduced in Williams, ‘Isidore’, 9. »
115     Paris, Bibliothèque nationale de France, Latin 8878, 45bisv–45ter. Reproduced in Williams, ‘Isidore’, 12. »
116     Vatican, Biblioteca Apostolica Vaticana, Vat. lat. 6018, 63v–64r; ed. Glorie, CCSL 175, 454–66; Gautier Dalché, MM, VatV17. Reproduced in Williams, ‘Isidore’, 16. On the Vatican Map, see Edson, Mapping Time and Space, 61–4, and Amat, ‘La Mappa mundi d’Albi’. »
117     CCSL 175, 458–9: ‘Ereptus Superior’, ‘mons Climax’ and ‘diserta Egypti’. »
118     CCSL 175, 462; cf. Orosius, 1.2.9 (ed. Zangemeister, 5). »
119     CCSL 175, 462–3. »
120     London, British Library, Cotton MS Tiberius B V, 56v; reproduced in Miller, kleineren, 33, and transcribed in Gautier Dalché, MM, Lon17, at 1:194: ‘Dara source’. On the Cotton Map, see Appleton, ‘Northern World’, and Alfred Hiatt, ‘An Anglo-Saxon world view? The Cotton mappa mundi Reassessed’, Viator 51, no. 2 (2020): 93–129. Orosius, 1.2.31 (ed. Zangemeister, 7). I interpret the second river as an extension of the Nile, due to its visual similarity and the interpretation of Orosius. On the Munich map, a separate Dara river (‘fluuius Dara’) is labelled flowing from Mount Atlas. »
121     Orosius, 1.2.9, ed. Zangemeister, 5; Gautier Dalché, MM, 1:194: ‘lacus Calearsum’. »
122     Orosius, 1.2.90, ed. Zangemeister, 13; cf. Isi., Etym. 9.2.15, 9.2.118–19; Gautier Dalché, MM, 1:194: ‘hic barbari Getuli’. »
123     London, British Library, Additional MS 10049, 64v; reproduced in P. D. A. Harvey, Medieval Maps of the Holy Land (London: British Library, 2012), 48–9. On the Tournai maps, previously called the ‘Jerome’ maps, see Harvey, Holy Land, 40–59, and Gautier Dalché, ‘Eucher de Lyon’, 6–9. »
124     Munich, Bayerische Staatsbibliothek, Clm 10058, 154v; Vatican, Biblioteca Apostolica Vaticana, Vat. lat. 6018, 63v–64r. »
125     London, British Library, Additional MS 10049, 64v: ‘India Egipti’. »
126     Pomponius Mela, De chorographia 3.96, ed. Silberman, 93; cf. Orosius, 1.2.31, ed. Zangemeister, 7, and Munich, Bayerische Staatsbibliothek, Clm 10058, 154v. »
127     London, British Library, Additional MS 10049, 64v: ‘Ethiopes Egiptii’. »
128     London, British Library, Additional MS 10049, 64v: ‘India Egipti’, ‘Ethiopes’ and ‘India Ethiopie’. »
129     Rouxpetel, L’Occident au miroir, 116–17; Gautier Dalché, ‘Eucher de Lyon’, 13. On the tripartition of India, see generally Marianne O’Doherty, The Indies and the Medieval West: Thought, Report, Imagination (Turnhout: Brepols, 2013), 33–8. »
130     Rouxpetel, L’Occident au miroir, 104–5, 117n64; von den Brincken, Nationes Christianorum orientalium, 246, 264–6, 338–9, 420–7. »
131     Harvey, Holy Land, 47–9. »
132     Cambridge, Corpus Christi College, MS 66, 2. »
133     Westrem, Hereford Map, 183, 185, my trans.: ‘Terminus Asye et Affrice’; ‘Terra Ethiopie’. Cf. Westrem’s discussion of the ambiguity of this region, which is also highlighted in the scribe’s confusion over Pliny’s measurement of Africa (Westrem, Hereford Map, 182, 340). »
134     Kugler, Ebstorfer Weltkarte, 1:96: ‘Mons Gathabathmon. In his montibus Affrica sumit initium’. »
135     Kugler, Ebstorfer Weltkarte, 1:68: ‘Egyptia deserta’. »
136     Kugler, Ebstorfer Weltkarte, 1:68: ‘In hoc latifundio et inter Egyptum et Ethyopiam et Libiam sunt genera simiarum silvestria …’ (In this region between Egypt, Ethiopia and Libya there are wild species of monkey …). Cf. C. Julius Solinus, Collectanea rerum memorabilium 27.55–60, ed. Theodor Mommsen (Berlin: Weidmann, 1895), 127–8. »
137     Kugler, Ebstorfer Weltkarte, 1:68: ‘Gentes Egyptie’. »
138     See Patrick Gautier Dalché, ‘La connaissance des régions septentrionales de l’Europe à l’époque carolingienne: un changement de point de vue’, Geographia Antiqua 25 (2016): 3–15; Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–71; Fraesdorff, Norden, 291–5; Wolever, ‘Post-Carolingian Geographies’, 153–63; and Thomas Foerster, ‘Imagining the Baltic: Mental Mapping the Works of Adam of Bremen and Saxo Grammaticus, Eleventh-Thirteenth Centuries’, in Imagining Communities on the Baltic Rim, from the Eleventh to Fifteenth Centuries, ed. Wojtek Jezierski and Lars Hermanson (Amsterdam: Amsterdam University Press, 2016), 37–58. »
139     Orosius, 1.2.54, ed. Zangemeister, 9; Isi., Etym. 14.4.3. »
140     Alheydis Plassmann, Origo gentis: Identitäts- und Legitimitätsstiftung in früh- und hochmittelalterlichen Herkunftserzählungen (Berlin: Akademie, 2006), 248–9; Leonid S. Chekin, Northern Eurasia in Medieval Cartography: Inventory, Text, Translation, and Commentary (Turnhout: Brepols, 2006), 22–5, 221, 248; Evelyn Edson, ‘Dacia ubi et Gothia: Die nordöstliche Grenze Europas in der mittelalterlichen Kartographie’, in Baumgärtner and Kugler, eds, Europa im Weltbild, 173–89; Wolever, ‘Post-Carolingian Geographies’, 153–63. See also the analogous issue around the relationship of Thule and Iceland: Dale Kedwards, The Mappae Mundi of Medieval Iceland (Cambridge: D. S. Brewer, 2020), 127–40. »
141     Alfred of Beverley, Historia 7, cf. 6, ed. J. P. T. Slevin and trans. L. Lockyer (Woodbridge: Boydell, 2023), 115, cf. 103: ‘iterum pagani de Danubia, Nowegia, Suethia, Gethia, Germania et aliis nationibus cum .viii. regibus et .xx. comitibus nauigio Britanniam aduecti’ (Once again pagans from the Danube, Norway, Sweden, Gothia, Germany and other nations arrived by ship in Britain, with eight kings and twenty earls). I have lightly edited Lockyer’s translation. The ultimate source for ‘Danubia’ is Asser’s De rebus gestis Aelfredi 21, ed. William Henry Stevenson, Asser’s Life of King Alfred (Oxford: Clarendon Press, 1959), 19, trans. Simon Keynes and Michael Lapidge, Alfred the Great: Asser’s Life of King Alfred and Other Contemporary Sources (London: Penguin, 1983), 74: ‘Et eodem anno magna paganorum classis de Danubia Britanniam advenit’ (In the same year, a great Viking fleet arrived in Britain from the Danube …). It is impossible to say whether this term is original to the ninth century or a corruption of the turn of the eleventh century introduced into the only (no longer) extant manuscript, but it cannot postdate this as it is already found in Byrhtferth of Ramsey, Historia Regum 4.6, ed. Michael Lapidge (Oxford: Clarendon Press, 2023), 132–3. As a result, this terms appears to be restricted to English chroniclers writing about the great pagan army, e.g. John of Worcester, Chronicon 888, ed. Reginald R. Darlington and Patrick McGurk, trans. Jennifer Bray and Patrick McGurk, The Chronicle of John of Worcester, vol. 2, The Annals from 450–1066 (Oxford: Clarendon Press, 1995), 280; Liber Eliensis 1.39, ed. E. O. Blake (London: Royal Historical Society, 1962), 53; and Roger of Howden, Chronica 866, ed. William Stubbs, 4 vols (London: Longmans, 1868–71), 1:38. This phenomenon was already identified by the Bollandists in their edition of the life of St. Æthelthryth, drawn from the Liber Eliensis, and subsequent scholars have generally shared their conclusion that this must be a corruption of a term for Danes or Denmark (Jean-Baptiste du Sollier, Jean Pien, Guillaume Cuypers and Petrus van den Bosch, eds, Acta Sanctorum, June, vol. 5 (Paris: Victor Palmé, 1867), 443C–D; cf. Stevenson in Asser, De rebus gestis Aelfredi, 217–18, and Keynes and Lapidge, trans., in Asser, De rebus gestis Aelfredi, 238). It bears noting here, however, that Eutropius (and thus Paul the Deacon and Landolf Sagax) describes Trajan’s foundation of a province across the Danube after his conquest of Dacia: ‘Daciam Decibalo victo subegit, provincia trans Danubium facta …’ (After defeating Decebalus he conquered Dacia and founded a province beyond the Danube). Eutropius, Breviarium ab urbe condita 8.2.2, ed. Carlo Santini (Leipzig: Teubner, 1979), 50; cf. Henry of Huntingdon, Historia Anglorum 1.25, ed. and trans. Diana Greenway (Oxford: Clarendon Press, 1996), 44–5. If this corruption does in fact date to the late tenth or early eleventh century, then this Danubian origin of the great pagan army offers a very interesting context for the near contemporary elision of Dacia and Dania by Dudo of Saint-Quentin (see p. 000 below). »
142     Patrick Gautier Dalché, ‘Comment penser l’Océan? Modes de connaissances des fines orbis terrarum du nord-ouest (de l’Antiquité au XIIIe siècle)’, in L’Europe et l’Océan au Moyen Age: Contribution à l’Histoire de la Navigation (Nantes: Cid, 1988), 217–20; e.g., Plato, Statesman 273d–e (trans. C. J. Rowe, in Plato: Complete Works, ed. John M. Cooper (Indianapolis: Hackett, 1997), 315). »
143     Vergil, Eclogae 1.66, ed. Silvia Ottaviano (Berlin: De Gruyter, 2013), 40, trans. Guy Lee (Harmondsworth: Penguin, 1984), 32–3: ‘et penitus toto divisos orbe Britannos’; cf. Horace, Odes 1.35.29–30, ed. D. R. Shackleton Bailey (Leipzig: Teubner, 1995), 38. See also Robert Coleman, Vergil Eclogues (Cambridge: Cambridge University Press, 1977), 86. »
144     Solinus, Collectanea rerum memorabilium 22.1, ed. Mommsen, 99–100: ‘Finis erat orbis ora Gallici litoris, nisi Brittania insula non qualibet amplitudine nomen paene orbis alterius mereretur’; Servius, In Vergili Bucolica, 1.66, ed. Thilo, 3:15. See also R. M. Thomson and M. Winterbottom, William of Malmesbury: Gesta Regum Anglorum, vol. 2, General Introduction and Commentary (Oxford: Clarendon Press, 1999), 48–9. For the use of this trope in relation to islands more generally, see Gautier Dalché, ‘Comment penser l’Océan’. »
145     Pliny, Naturalis historia 4.16.102–4, ed. Rackham, 2:196–9; Pomponius Mela, De chorographia 3.49–58, ed. Silberman, 81–3; cf. Strabo, Geography 4.5, trans. Jones, 2:253–63. On the early history of Thule, see Kedwards, Medieval Iceland, 129–32. »
146     Martianus Capella, De nuptiis 6.666, ed. Willis, 235; Bede, Historia ecclesiastica 1.1, ed. Colgrave and Mynors, 14–15. »
147     Isi., Etym. 14.6.2–6, trans. Stephen A. Barney et al. (Cambridge: Cambridge University Press, 2006), 294: ‘Thyle ultima insula Oceani inter septentrionalem et occidentalem plagam ultra Brittaniam’. »
148     Orosius, 1.2.76–80, ed. Zangemeister, 12. »
149     Ps.-Aeth., 1.16, ed. Riese, 78; Jul. Hon., Cosm. 14, ed. Riese, 33: ‘Sea of Thule’. »
150     Jordanes, Getica 1.8, ed. Theodor Mommsen, MGH AA 5.1 (Berlin: Weidmann, 1882), 55–6, trans. Peter Van Nuffelen and Lieve Van Hoof (Liverpool: Liverpool University Press, 2020), 224–5: ‘Habet et in ultimo plagae occidentalis aliam insulam nomine Thyle …’ ([The ocean] holds also, in its most western region, another island called Thule …). For an overview of recent literature on the relationship of Thule and Scanza in Jordanes, see Van Nuffelen and Van Hoof, trans., in Jordanes, Getica, 44–8, esp. 48n225. On Jordanes’s use of Julius Honorius, see Nicolet and Gautier Dalché, ‘Les “quatre sages”’, 163, 188. »
151     Jordanes, Getica 2.10, MGH AA 5.1, 56. »
152     Gildas, De exidio 3.1, ed. and trans. Michael Winterbottom (London: Phillimore, 1978), 89, 16: ‘Brittannia insula in extremo ferme orbis limite circium occidentemque’. Cf. John Morris in Gildas, De exidio, 148. »
153     HoSV, Desc. prol., ed. Gautier Dalché, 133. »
154     HoSV, Desc. 26, ed. Gautier Dalché, 159. Cf. Orosius, 1.2.76, ed. Zangemeister, 12. »
155     HoSV, Desc. 25, ed. Gautier Dalché, 157; Hugh of Fleury, Historia ecclesiastica 5.prol., ed. Leendert Martin de Ruiter (PhD diss., Rijksuniersiteit Gronigen, 2016), 103; cf. Orosius, 1.2.63, ed. Zangemeister, 11, trans. Fear, 43: ‘Gallia Belgica habet … a circio oceanum Britannicum, a septentrione Britanniam insulam’ (Gallia Belgica has … to the north-west, the British Ocean; and to the north, the island of Britain). The italicised section has been removed in the Hughs’ description. »
156     Paris, Bibliothèque nationale de France, Latin 15009, 17r: ‘In oceano septemtrionali ab oriente ad occidentem’. »
157     On the different ordering of islands in Lambert and Isidore, see Gautier Dalché, ‘Maps in Words’, 229; Patrick Gautier Dalché, ‘De la gloss à la contemplation: Place et fonction de la carte dans le manuscrit du haut Moyen Age’, in Géographie et culture, 746–8; and Derolez, The Autograph Manuscript, 74–5. »
158     For the cartographical connection, see Gautier Dalché, ‘Maps in Words’, 228–9; Lamb., Lib. flor., 51v: ‘Brittannia contra galliam in oceano … Thile ultra brittanniam … Hibernia proxima brittannie’. »
159     Lamb., Lib. flor., 51v: ‘Hiberus inter hispaniam et brittaniam … Orcades XXXIII inter britanniam et gothiam’. »
160     Lamb., Lib. flor., 241r; Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf. 1 Gud. lat., 69v–70r (Gautier Dalché, MM, Wolf13). »
161     Lamb., Lib. flor., 234r–241r; on the contents of this section and its complicated history of compilation, see Derolez, The Autograph Manuscript, 164–6. If my suggestion about the map’s relevance to Frankish origins is correct, then this might explain its unusual placement highlighted by Derolez (The Autograph Manuscript, 152). »
162     Hon., Im. mun. 1.29, 33–5, ed. Flint, 62–6: ‘in Europa’. »
163     Hon., Im. mun. 1.29, ed. Flint, 62: ‘Contra Hispaniam versus occasum sunt in oceano hee insule … Ultra hanc versus aquilonem est mare congelatum et frigus perpetuum’. »
164     See Flint in Hon., Im. mun. 46–7, passim, and p. 000 above. »
165     Gildas, De exidio 3.1, ed. Winterbottom, 89. »
166     Ireland also sometimes depicted clockwise to the north, as on Lambert’s maps. »
167     Guido of Pisa, Geographica 130, ed. Schnetz, 142: ‘Hispania habet fines ab uno latere iuxta litus occeani saltum Pireneum, Spanosguasconiam vero iuxta mare magnum Gallicum, ipsumque ex alio latere saltum Pirineum et provinciam Septimanam. a secundo quoque latere habet fines Hispania saepe dictum mare Gallicum Valeriacum. a tertio autem id est ab occidentali quod praediximus mare occeano Britannicum. a quarto demum latere habet fines praemissum fretum Gaditanum, quod exit de mari Gallico et ingreditur in praedictum occeanum occidentalem Britannicum, quodque dividit inter Africae Maurorum patriam et eandem Hispaniam’. Cf. Ravennatis anonymi cosmographia 4.45–6, ed. Schnetz, 82–3. I have identified the rather puzzling ‘Valeriacum’ according to its use by the Ravenna Anonymous, who Guido follows here. That work confuses a variety of letters in classical orthography, such as v and b, making the ‘Baliares’ of Pliny ‘Valeares’, which are then situated in the ‘mare Gallicum Valeriacum’, itself presumably a derivative of Pliny’s classical ‘mare Baliaricum’ (see Louis Dillemann, La Cosmographie du Ravennate (Brussels: Latomus, 1997), 19, 191–2). Cf. Pliny, Naturalis historia 3.5.74–8, ed. Rackham, 2:56–7, and Ravennatis anonymi cosmographia 5.27, ed. Schnetz, 102–3. »
168     Ravennatis anonymi cosmographia 5.31–2, ed. Schnetz, 108; see also Dillemann, La Cosmographie, 195–6. »
169     Oxford, St. John’s College, MS 17, 6r; Gautier Dalché, MM, Oxf23. On this map, see Edson, Mapping Time and Space, 86–95. »
170     Reykjavík, Stofnun Árna Magnússonar, GkS 1812 III 4to, 5v–6r; reproduction, transcription and commentary in Kedwards, Medieval Iceland, 197–229, on this map: 101–18. »
171     Kedwards, Medieval Iceland, 112. »
172     Alfred Hiatt, ‘From Pliny to Brexit: Spatial Representation of the British Isles’, Postmedieval: A Journal of Medieval Cultural Studies 7, no. 4 (2016): 516–19. »
173     Dublin, National Library of Ireland, MS 700, 48r; Gerald of Wales, Topographia Hibernica 1, ed. John J. O’Meara, ‘Giraldus Cambrensis in Topographia Hibernie. “Text of the First Recension”’, Proceedings of the Royal Irish Academy: Archaeology, Culture, History, Literature 52 (1948–50): 120: ‘Insularum occidentalium hec ultima [sc. Hibernia] … habet; Britanniam maiorem ab oriente; solum occeanum ab occidente’ (This last of the western islands [sc. Ireland] has Great Britain to the east and only ocean to the west). See also Thomas O’Loughlin, ‘An Early Thirteenth-Century Map in Dublin: A Window into the World of Giraldus Cambriensis’, Imago Mundi 51, no. 1 (1999): 24–39. »
174     Gautier Dalché, ‘Introduction’, 8–11; Gautier Dalché, ‘Considérations intempestives’. »
175     Stephen McKenzie, ‘The Westward Progression of History on Medieval Mappaemundi’, in Harvey, ed., Hereford World Map, 335–44, highlights many of the issues around ‘reading’ maps. Gautier Dalché, ‘Introduction’, 10, also notes some of the more egregious examples. »
176     See Mauntel, Erdteile, 257–68. See also Natalia I. Petrovskaia, ‘La formulation tripartite du monde dans les encyclopédies et textes littéraires du Moyen Âge’, in La Formule au Moyen Âge IV, ed. Élise Louviot, Charles Garcia and Stephen Morrison (Turnhout: Brepols, 2021), 197–217, whose theory that the two orderings (As, E, Af and As, Af, E) can be divided respectively into an Isidorian and Orosian tradition strikes me as unconvincing; on Petrovskaia’s application of the Translatio imperii, cf. Wolever, ‘Translatio imperii’. »
177     Oschema, Bilder von Europa, 206–8. »
178     Hon., Im. mun. 1.7, ed. Flint, 52; Lamb., Lib. flor., 19r. »
179     HoSV, Desc. 7, ed. Gautier Dalché, 138–9. »
180     Gautier Dalché, ‘Maps in Words’. »
181     Mauntel, Erdteile, 155–65. »
182     Mauntel, Erdteile, 101–3; Salway, ‘Putting the World in Order’, 213–16; Ingrid Baumgärtner, ‘Europa in der Kartographie des Mittelalters: Repräsentationen – Grenzen – Paradigmen’, in Baumgärtner and Kugler, eds, Europa im Weltbild, 11–13; Oschema, Bilder von Europa, 89–90. »
183     Libya and Africa have been amalgamated under ‘Africa’. E, Af, As: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Lucan, 6.817, ed. Shackleton Bailey, 162; Isi., Etym. 13.16.1. E, As, Af: Pliny, Naturalis historia 3.1.3, ed. Rackham, 2:4–5; Cicero, De natura deorum 2.165, ed. W. Ax (Berlin: De Gruyter, 2008), 116; Strabo, Geography 1.4.7, trans. Jones, 1:245; cf. Martianus Capella, De nuptiis 6.622, ed. Willis, 217. Some Christian authors also used these orderings, e.g. Boethius, De diuisione, PL 64, 888C; Cassiodorus, Historia tripartita 3.3.13, ed. W. Jacob and M. Adriaen, CSEL 71 (Vienna: Tempsky, 1952), 141; and Isidore of Seville, De natura rerum 48.2, ed. Fontaine, 324–5. Finally, these orderings were transmitted via Pliny to some early medieval authors: Bede, De natura rerum 51, CCSL 123A, 233–4; Bede, Expositio actuum apostolorum 10.7, ed. M. L. W. Laistner, CCSL 121 (Turnhout: Brepols, 1983), 49; and Dicuil, Liber de mensura orbis terrae prol.6, 1–3, ed. Tierney, 44–55. »
184     Augustine, De civitate Dei 16.17, CCSL 48, 521; Orosius, 1.2.1–11, 6.17.4, ed. Zangemeister, 5, 218; Quoduultdeus, Liber promissionum, Dimidium temporis 11.19, ed. R. Braun, CCSL 60 (Turnhout: Brepols, 1976), 204; Jerome, In Hiezechielem 2.5.5–6, CCSL 75, 56; Jerome, In Dan. 1.4.7b–8, CCSL 75A, 813; Jerome, In epistulas Paulinas ad Titum 1.12–14, ed. F. Bucchi, CCSL 77C (Turnhout: Brepols, 2003), 30; Cassiodorus, Institutiones 2.3.10, ed. Mynors, 114; Isi., Etym. 14.2.1; Ps.-Aeth., 1.1, ed. Riese, 71; Jordanes, Getica 1.4, MGH AA 5.1, 54; Bede, Homeliae evangelii 2.13, ed. D. Hurst, CCSL 122 (Turnhout: Brepols, 1955), 269. It is worth noting the prominence of Sallust, Bellum Iugurthinum 17.3, ed. Kurfess, 68, which lists Asia and Europe in that order after noting that Africa is normally thought of as the world’s ‘third part’. »
185     There is one off-hand occurrence of an As, Af, E ordering in Orosius, 6.1.6, ed. Zangemeister, 187; cf. Petrovskaia, ‘La formulation tripartite’, 206. »
186     See generally Lozovsky, Earth Is Our Book. There is somewhat more for late antiquity, e.g. Merrills, History and Geography, and Gautier Dalché, ‘L’enseignement de la géographie’. »
187     Gautier Dalché, ‘L’enseignement de la géographie’, 163–74, and Patrick Gautier Dalché, ‘Cartes et enseignement de la géographie durant le haut Moyen Âge: L’exemple d’un manuel inédit’, in Du copiste au collectionneur. Mélanges d’histoire des textes et des bibliothèques en l’honneur d’André Vernet, ed. Donatella Nebbiai-Dalla Guarda and Jean-François Genest (Turnhout: Brepols, 1999), 54; Lozovsky, Earth Is Our Book, ch. 4, esp. 138. »
188     Situs orbis 5.1, ed. Gautier Dalché, 164; Descriptio terrarum, Albi, Bibliothèque municipale, MS 29, 58v; Versus de Asia et de uniuersi mundi rota, ed. Glorie, CCSL 175, 441–54; Commentaria Bernensia 9.411, ed. Hermann Usener, Scholia in Lucani Bellum civile: Pars prior (Leipzig: Teubner, 1869), 301; Rabanus Maurus, De universo 12.2, PL 111, 333C–D. »
189     Otto of Freising, Chronica 1.1, MGH SS rer. Germ. 45, 37; cf. William of Conches, Dragmaticon 6.5, CCCM 152, 193; William of Conches, Glossulae super Boethium 3.m12, ed. L. Nauta, CCCM 158 (Turnhout: Brepols, 1999), 208; Sigebert of Gembloux, Chronica 738, ed. Ludwig Bethmann, MGH SS 6 (Hannover: Hahn, 1844), 331 [hereafter Sig., Chron.]. »
190     London, British Library, Cotton MS Tiberius B V, 57r, 60v: ‘huic operi de tribus partibus, videlicet Asia, Africa, Europa, mappam depinxerat aptam’; transcription in Gautier Dalché, ‘L’héritage antique’, 46n41. On the relationship of the incipit to the map, see Alfred Hiatt, ‘The Cotton Mappa mundi as analogue to the Albi world map’, in Vagnon and Victor, eds, La Mappa mundi d’Albi, 199–201. »
191     Priscian, Periegesis 15, ed. van de Woestijne, 38. »
192     Gn 5:31, 6:10, 7:13, 9:18, 10:1; Eucher of Lyon, Instructionum ad Salonium 1.19, ed. C. Mandolfo, CCSL 66 (Turnhout: Brepols, 2003), 87; Bede, In principium Genesis 3.10.1–2, ed. C. W. Jones, CCSL 118A (Turnhout: Brepols, 1967), 142. Both the Liber de mundi institucione, ed. Michel Zimmerman, ‘Le monde d’un catalan au Xe siècle: Analyse d’une compilation isidorienne’, in Le métier d’historien au moyen-âge: Etudes sur l’historiographie médiévale, ed. Bernard Guenée (Paris: Publications de la Sorbonne, 1977), 74–8, and Rabanus Maurus, De universo 2.1, PL 111, 34D associate the three parts with Noah’s sons, but they nevertheless follow the As, E, Af ordering in their geographical description. »
193     Paris, Bibliothèque nationale de France, Latin 8669, 77r; transcription in Lozovsky, Earth Is Our Book, 134n109, cf. 116. Cf. Heiric of Auxerre, Homiliae per circulum anni, pars aestiua 9, ed. R. Quardi, CCCM 116B (Turnhout: Brepols, 1994), 71; Glossae collectae e primo secundoque libris Orosii historiarum contra paganos 1.2.95, ed. Olivier Szernwiniack, ‘Un commentaire hiberno-latin des deux premiers livres d’Orose, Histoires contre les paines’, Archivum latinitatis medii aevi 51 (1992/3): 70–1, where the comment on the three parts is a deviation from the section of Isidore being copied. A French origin would also tally well with Richer of Saint-Remi, Historiarum libri IIII 1.1, ed. Hartmut Hoffmann, MGH SS 38 (Hanover: Hahn, 2000), 36, but examples of an As, Af, E ordering are already found in Italy by the mid-eleventh century (e.g. Bruno of Segni, Commentaria in Ioannem 3.54, PL 165, 599B; cf. Mauntel, Erdteile, 261n1118) and in Salzburg by the twelfth century, where the first scribe of a Waltharius manuscript has written ‘Asia, Affica, Europa’ twice as penmanship practice on the preceding folio, perhaps following the poem’s opening line: ‘Tertia pars orbis, fratres, Europa vocatur’ (The third part of the world, brothers, is called Europe). Vienna, Österreichische Nationalbibliothek, Cod. 289, 103r; Waltharius 1, ed. Karl Strecker, MGH Poetae 6.1 (Weimar: Hermann Böhlaus, 1951), 24. »
194     The Expositio mappe mundi by Roger of Howden (?), ed. Patrick Gautier Dalché, Du Yorkshire a l’Inde: Un ‘géographie’ urbaine et maritime de la fin du xiie siècle (Roger de Howden?) (Geneva: Droz, 2005), 143–64, also follows the order As, Af, E. See generally, Gautier Dalché, ‘Comment et pourquoi décrire une mappemonde’. »
195     Gautier Dalché, ‘Maps in Words’, 227–34; Terkla, ‘Hugh of Saint-Victor’. »
196     Paris, Bibliothèque nationale de France, Latin 15009, 16r–v. »
197     Lamb., Lib. flor., 48r–50r. »
198     Lamb., Lib. flor., 51v. »
199     Lamb., Lib. flor., 52v. »
200     Isi., Etym. 14.3–5; Ps.-Aeth., 1.7–12, 20–5, 33–7, ed. Riese, 74–7, 81–8, 86–7; cf. Gautier Dalché, ‘Maps in Words’, 228–9; Derolez, The Autograph Manuscript, 67, 73–5. »
201     Isi., Etym. 14.6.2–13. Derolez suggests the source of the islands is simply Isidore, though Gautier Dalché highlights various differences between Lambert and Isidore, which he plausibly attributes to Lambert’s use of a map to aid his activity of compilation. Derolez, The Autograph Manuscript, 74n108; Gautier Dalché, ‘Maps in Words’, 228–9. »
202     Though the grand mappa mundi from the autograph manuscript has been lost, there is a version from a late twelfth-century copy in Wolfenbüttel, Herzog August Bibliothek, Cod. Guelf. 1 Gud. lat., 69v–70r, which is normally taken to be accurate. Danielle Lecoq, ‘La Mappemonde du Liber Floridus ou La Vision du Monde de Lamber de Saint-Omer’, Imago Mundi 39 (1987): 9–49; Gautier Dalché, ‘De la glose’, 746–7; Gautier Dalché, ‘Maps in Words’, 227–31. »
203     Gautier Dalché, ‘De la glose’, 746–9. »
204     Gautier Dalché, ‘Maps in Words’, 230–1. »
205     Isi., Etym. 14.2.2: ‘Nam Asia a meridie per orientem usque ad septentrionem pervenit; Europa vero a septentrione usque ad occidentem; atque inde Africa ab occidente usque ad meridiem’. I have modified the translation of Barney et al., 285. Cf. Augustine, De civitate Dei 16.17, CCSL 48, 521. »
206     Hon., Im. mun. 1.7, ed. Flint, 52: ‘Asia a septentrione per orientem usque ad meridiem, Europa ab occidente usque ad septentrionem, Affrica a meridie usque ad occidentem extenditur’. »
207     Both begin their discussion of Europe with Scythia in the north-east and end with Spain in the south-west and they begin their discussion of Africa with Libya in the north-east and end with Ethiopia in the south. »
208     Honorius Augustodunensis, De exsilio et patria animae, PL 172, 1244D: ‘In hac Aratus mappam mundi expandit, in qua Asiam, Africam, Europam ostendit’ (In this [city of Geometry], Aratus lays out a world map, on which Asia, Africa and Europe are shown …); cf. Gautier Dalché, ‘Maps in Words’, 230n28. »
209     Gautier Dalché, ‘Maps in Words’, 236. »
210     Isi., Etym. 14.3.2, trans. Barney et al., 285: ‘Paradisus est locus in orientis partibus constitutus’; Hon., Im. mun. 1.8, ed. Flint, 52: ‘Huius prima regio in oriente est paradysus’. »
211     Hon., Im. mun. 1.10–17, at 17, ed. Flint, 53–7: ‘Hee superius dicte regiones ab oriente incipientes, recta linea ad Mediterraneum mare extenduntur’. »
212     Hon., Im. mun. 1.18, ed. Flint, 58: ‘Suprascriptis regionibus versus aquilonem annectuntur he regiones’ (These regions adjoin the previously discussed regions to the north). »
213     Hon., Im. mun. 1.30, ed. Flint, 63: ‘Hec in oriente Indi fluminis surgit, et per meridiem vergens, in occidentem tendit’. »
214     Hon., Im. mun. 1.32, ed. Flint, 64: ‘Versus meridiem vero est Ethiopia …’. »
215     Hon., Im. mun. 1.32, ed. Flint, 64: ‘In extremis finibus Africe versus occidentem …’. »
216     Hon., Im. mun. 1.22, ed. Flint, 59: ‘A Tanai fluvio … versus meridiem usque ad Danubium porrigitur’. »
217     Hon., Im. mun. 1.25–9, ed. Flint, 60–3. »
218     Gautier Dalché, ‘Représentations géographiques de l’Europe’, 69–73; Fraesdorff, Norden, ch. 3; Klaus Oschema, ‘Der Europa-Begriff im Hoch- und Spätmittelalter. Zwischen geographischem Weltbild und kultureller Konnotation’, Jahrbuch für europäische Geschichte 2 (2001): 196–8. »
219     Cf. Oschema, ‘Europa-Begriff’, 197. »
220     Mauntel, Erdteile, 141–52, 155–65. »
221     Paris, Bibliothèque nationale de France, Latin 15009, 16r (emphasis added): ‘In latere asie ad aquilonem ab occeano orientali occidentem uersus porrigitur / [s]cythia’; ‘In latere australi asie porrigitur mare ruberum ab oceano orientali occidentem uersus usque ad confinia arabie et palestine’ (On the side of Asia to the north, Scythia extends from the eastern ocean towards the west; On the southern side of Asia, the Red Sea extends from the eastern ocean towards the west up to the boundaries of Arabia and Palestine). »
222     Paris, Bibliothèque nationale de France, Latin 15009, 16r–v. »
223     Richard of Saint Victor, Liber Exceptionum 1.3.3, cf. 1.3.1, 4, passim, ed. Jean Châtillon (Paris: Vrin, 1958), 124, cf. 122, passim: ‘Affrice regiones ab Oriente ad Occidentem porriguntur sic …’. »
224     Hon., Im. mun. 2.3, ed. Flint, 92–3; on some of the wider historical resonances, see Wolever, ‘Translatio imperii’. »
225     Recent scholarship has increasingly highlighted the fundamental plurality of models in medieval geography: Akbari, Idols in the East, 24–50; Baumgärtner, ‘Winds and Continents’; cf. Mauntel, Erdteile, 19, and Nathalie Bouloux, ‘Ordering and Reading the World: The Maps in Lambert of Saint-Omer’s “Liber Floridus”’, in Mauntel, ed., Geography and Religious Knowledge, 97–9. »
226     ‘Sallust map’ simply refers to any map or T-O diagram included in a manuscript of Sallust’s Bellum Iugurthinum, typically near 17.3, ed. Kurfess, 68. On these maps, see generally Patrick Gautier Dalché, ‘Les cartes des manuscrits de Salluste’, Geographia Antiqua 30 (2021): 177–96; Mauntel, Erdteile, 393–8; Edson, Mapping Time and Space, 18–21; David Woodward, ‘Medieval Mappaemundi’, in Harley and Woodward, eds, History of Cartography, 343–4; and Miller, kleineren, 110–15. On diagrams in classical works, see more broadly Patrick Gautier Dalché, ‘Les diagrammes topographiques dans les manuscrits des classiques latins (Lucain, Solin, Salluste)’, in La tradition vive. Mélanges d’histoire des textes en l’honneur de Louis Holtz, ed. P. Lardet (Turnhout: Brepols, 2003), 291–306. »
227     Paris, Bibliothèque nationale de France, Latin 6089, 26r; Cambridge, Gonville and Caius College, MS 719, 37v; Milan, Biblioteca Ambrosiana, G 21 Sup., 23v; Venice, Biblioteca Nazionale Marciana, Fondo Ant. Lat. Z 432 (1656), 40r. See respectively Paris: Gautier Dalché, MM, Par15 (C12 2/2, Italy?) and Burger Munk Olsen, Catalogue des manuscrits classiques latins copiés du IXe au XIIe siècle, 5 vols (Paris: CNRS, 1982–2020), 2:343 (C12 2/2, Italy); Cambridge: M. R. James, A Descriptive Catalogue of the Manuscripts in the Library of Gonville and Caius College (Cambridge: Cambridge University Press, 1908), 690–1 (C13, Italy); Milan: Robert Black, ‘The Rise and Fall of the Latin Classics’, Aevum 91, no. 2 (2017): 432 (C12 2/2, Italy; but, 432n147, previously dated C13ex, Venice); Venice: Anna Novokhatko, ‘Eine Liste der Handschriften der im Sallust- und Cicerocorpus überlieferten Invektiven (Sallusti in Ciceronem et invicem Invectivae)’, Eikasmos 13 (2002): 281 (citing Zorzanello’s catalogue C14). The Paris, Cambridge and Milan manuscripts are reproduced in Chekin, Northern Eurasia, 341, 343, 346; likewise, the Paris, Cambridge and Venice manuscripts in Woodward, ‘Mappaemundi’, fig. 18.45, 18.50, 18.74. »
228     These features are also found on other Sallust maps, e.g., Paris, Bibliothèque nationale de France, Latin 6253, 52v (Gautier Dalché, MM, Par16). »
229     The major variants are East: ortum ] ortorum V M; South: caritatis significat deest C; West: uitiorum ] uiuorum V M, occasum ] occasus C P, significat ] designat M dicitur C; North: constrictionem ] districtionem P, uoluptatis ] uoluntatis V M, et deest V M. ‘the east signifies the light of faith or the rise of good works’; ‘the south signifies the clarity of wisdom and the warmth of love’; ‘the west signifies the fall of vices and sins’; ‘the north signifies the binding of delight and mortification [of desires]’. ‘Of desires’ is missing on the Sallust maps, but found in Rabanus Maurus, Commentariorum in Genesim 3.14, PL 107, 593B–C: ‘Occidens in hoc loco occasum vitiorum et peccatorum designat, Oriens lucem fidei vel ortum bonorum operum, Septentrio constrictionem voluptatis et mortificationem desideriorum. Meridies claritatem sapientiae et fervorem charitatis non inconvenienter ostendit’ (See the above trans. in the order: west, east, north, south). »
230     V also contains a selection of Cicero’s orations and a pseudonymous exchange of invectives between Cicero and Sallust. »
231     See generally Walter Berschin, ‘Eine griechisch-althochdeutsche-lateinische Windrose von Froumund von Tegernsee im Berlin-Krakauer Codex lat. 4°939’, in Vetustatis amore et studio: Księga pamiątkowa ofiarowana Profesorowi Kazimierzowi Limanowi, ed. Ignacego Lewandowski and Andrzeja Wójcika (Poznań: Uniwersytet Adama Mickiewicza, 1995), 23–30 (repr. in Walter Berschin, Mittellateinische Studien, vol. 1 (Heidelberg: Mattes Verlag, 2005), 285–92). »
232     Kraków, Biblioteka Jagiellońska, Nr. 834 (previously Berlin, Staatsbibliothek Prußischer Kulturbesitz, Ms. lat. 4° 939). See Rolf Bermann et al., eds, Katalog der althochdeutschen und altsächsischen Glossenhandschriften (Berlin: De Gruyter, 2005), nr. 45. The database is accessible online at glossen.germ-ling.uni-bamberg.de (accessed September 10, 2023). »
233     The subscription is transcribed in Die Tegernseer Briefsammlung (Froumund), ed. Karl Strecker (Berlin: Weidmann, 1925), XI. »
234     The Old High German is based on Einhard, Vita Karoli 29, ed. G. H. Pertz, G. Waitz and O. Holder-Egger, MGH SS rer. Germ. 25 (Hannover: Hahn, 1911), 34. Paulo Farmhouse Alberto, ‘The Textual Tradition of the Carmen de uentis (AL 484): Some Preliminary Conclusions with a New Edition’, Aevum 83 (2009): 341–75. »
235     On ADAM and the cardinal points, see, e.g., Augustine, Enarrationes 95.15, CCSL 39, 1352–3, and generally Kochanek, Vorstellung vom Norden, 164–218, esp. 208–13. »
236     Berschin, ‘Windrose’, 28: ‘Oriens lucem fidei vel ortum bonorum operum significans. … – Meridies claritatem sapientie et fervorem caritatis ostendens. … – Occidens vitiorum et peccatorum occasum designans. … – Septentrio constrictionem voluptatis et mortificationem desideriorum exprimit. …’ (see trans. as in Rabanus at p. 000 above). Notable differences between the windrose (WR), Sallust map (TO) and Rabanus (RM) include, Oriens: vel WR RM ] et TO; Meridies: ostendens WR ] ostendit RM significat TO; Occidens: designans WR ] designat RM significat TO; Septentrio: desiderorum WR RM deest TO. »
237     The wind diagram is found in London, British Library, Harley MS 2688, 17r, which was originally part of the first quire of London, British Library, Harley MS 3095. On the dating and location of these manuscripts, see Bernard Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts, 3 vols (Wiesbaden: Harrassowitz, 1998–2014), 2:114, 121, and Sigrid Krämer, Handschriftenerbe des deutschen Mittelalters, 3 vols (Munich: C. H. Beck, 1989–90), 2:417. Bischoff dates the former quire C9 4/4, but expresses some doubts about placing the manuscript in Cologne, suggesting France(?) instead and dates the latter C10 1/2, which Krämer places confidently in the possession of Cologne Cathedral. Froumund’s use of this diagram in his copy of Boethius provides a terminus ante quam around the 990s for the combined manuscript’s presence in Cologne. »
238     Georg Schepss, Handschriftliche Studien zu Boethius De consolatione philosophiae (Würzburg: Thein, 1881), 3. »
239     Paris, Bibliothèque nationale de France, Latin 10195, 1r. See Claudine Moulin and Elvira Glaser, ‘Echternacher Glossenhandschriften’, in Die althochdeutsche und altsächsische Glossographie: Ein Handbuch, ed. Rolf Bergmann and Stefanie Stricker (Berlin: De Gruyter, 2009), 1267–8. »
240     Munich, Bayerische Staatsbibliothek, Clm 15825, 1r. »
241     Dijon, Bibliothèque municipale, MS 448, 75r. On this manuscript, see Colette Jeudy and Yves François Riou, Les manuscrits classiques latins des bibliothèques de France, vol. 1 (Paris: CNRS, 1989), 483–500. »
242     The Kraków, London, Paris and Munich manuscripts all have ‘trahuntur’ instead of ‘iugantur’, whereas the Dijon manuscript reads ‘ligantur’. The Dijon manuscript also breaks with the other four and follows the beta text by replacing ‘hos circum’ with ‘hinc quoque sex’; cf. Alberto ‘Carmen de uentis’, 360, 365–9, 374. »
243     Vatican, Biblioteca Apostolica Vaticana, Pal. lat. 1356, 113r; Gautier Dalché, MM, VatP14, and Ludwig Schuba, Die Quadriviums-Handschriften der Codices Palatini Latini in der Vatikanischen Bibliothek (Wiesbaden: Dr. Ludwig Reichert Verlag, 1992), 34–9. See Gautier Dalché, ‘Les cartes des manuscrits de Salluste’, 184–9. »
244     Vatican, Biblioteca Apostolica Vaticana, Pal. lat. 1356, 113r; Gautier Dalché, MM, 1:404: ‘Oriens: lucem fidei ortum bonorum operum figurat’; ‘Meridies: claritatem sapientie fervorem caritatis figurat’; ‘Occidens: viciorum peccatorum [occasum] figurat’; ‘Septentrio: constrictionem voluptatis et mortificationem’ (see trans. as in the Sallust maps at p. 000 above). »
245     See esp. Edson, Mapping Time and Space, 36–96; cf. Kedwards, Medieval Iceland, 63–100. »
246     Woodward, ‘Mappaemundi’, 335–7; Bronder, ‘Das Bild der Schöpfung’; Kochanek, Vorstellung vom Norden, 164–218; Wolever, ‘Post-Carolingian Geographies’, 163–7. »
247     Baumgärtner, ‘Winds and Continents’, 54–6; Thomas Raff, ‘Die Ikonographie der mittelalterlichen Windpersonifikationen’, Aachener Kunstblätter 48 (1978/9): 140–55. »
248     Obrist, ‘Wind Diagrams’, 34, 70–84. Cf. Gautier Dalché, ‘Considérations intempestives’. »
249     Obrist, ‘Wind Diagrams’, 45, and fig. 4 (Munich, Bayerische Staatsbibliothek, Clm 16128, 35v). »
250     Obrist, ‘Wind Diagrams’, 45–9, and, e.g., fig. 3 (Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 2077, 99r) and fig. 7 (Bern, Burgerbibliothek, Cod. A 92.20, 2r). »
251     Obrist, ‘Wind Diagrams’, 49–53, and figs 9 and 10 (Bern, Burgerbibliothek, Cod. 212, 109r, Gautier Dalché, MM, Bern8; Berlin, Staatsbibliothek, MS Phill. 1830, 3v, Gautier Dalché, MM, Berl21). »
252     Obrist, ‘Wind Diagrams’, 47, and figs 3 and 8, cf. 19 (Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 2077, 99r; Basel, Universitätsbibliothek, MS F III 15a, 22r; cf. Cologne, Erzbischöfliche Diözesan- und Dombibliothek, MS 83 II, 84r). »
253     Obrist, ‘Wind Diagrams’, 52, and fig. 10 (Berlin, Staatsbibliothek, MS Phill. 1830, 3v); see also Karlsruhe, Badische Landesbibliothek, Cod. Aug. perg. 106, 52v (Gautier Dalché, MM, Kar2); Vatican, Biblioteca Apostolica Vaticana, Reg. lat. 1260, 39r (Gautier Dalché, MM, VatR20). »
254     Paris, Bibliothèque nationale de France, Latin 5239, 139r; Strasbourg, Bibliothéque nationale et universitaire, MS 326, 120v (Gautier Dalché, MM, Stra2). See Obrist, ‘Wind Diagrams’, 52. »
255     Cf. Obrist, ‘Wind Diagrams’, 52n89. »
256     Strasbourg, Bibliothéque nationale et universitaire, MS 326, 120v: ‘hic oriens oritur’. Similarly in Paris, Bibliothèque nationale de France, Latin 13953, 72v (Gautier Dalché, MM, Par72). »
257     Paris, Bibliothèque nationale de France, Latin 10837, 42r. On this illustration, with a transcription, see Barbara Obrist, ‘The Astronomical Sundial in Saint Willibrord’s Calendar and its Early Medieval Context’, Archives d’histoire doctrinale et littéraire du Moyen Age 67 (2000): 71–118. »
258     Obrist, ‘Astronomical Sundial’, 76–8, notes that one is Latin and another Greek, but cannot identify the third. »
259     Obrist, ‘Astronomical Sundial’, 77: ‘[Aquilo] significat antichristum mundum stringentem ad senectutem’. »
260     Paris, Bibliothèque nationale de France, Latin 10195, 33v and 57r. »
261     Milan, Biblioteca Ambrosiana, G 21 Sup. See Black, ‘Rise and Fall of the Latin Classics’, 432. »