Published on August 31, 2025

Home » Historical Maps of Brazil and the World » The Catalan Atlas: Redrawing the Medieval World and Asia

Catalan World Maps: Transforming Medieval Cartography and the Representation of Asia

Introduction

Another notable stage was reached in the fourteenth century, when European cartographers made the first attempt since classical times to include the continent of Asia within their world picture on the basis of contemporary knowledge.

The results of these efforts are embodied in the series of Catalan world maps.

Catalan school and Majorca

In the first half of the fourteenth century the Catalan school, largely Majorcan, took over from the north Italians the lead in cartographic progress, though rather as successors than as innovators.

During the preceding century the Majorcans had earned a great reputation among the peoples of the western Mediterranean for their maritime prowess.

After their incorporation in the Aragonese confederation (1229), the three ports of Palma, Barcelona, and Valentia formed the basis of a commercial enterprise which extended to most of the north African ports as far as Egypt, and beyond to Syria.

Early in the century the population had been augmented by Jewish refugees from the Almohad persecutions, and this element strengthened commercial relations especially with Morocco.

Trade was also stimulated by the aggressive policy of the able rulers of Aragon, and diplomatic agents appear by 1300 to have reached as far afield as Persia.

These Jewish refugees also included scholars who could interpret the works of Arab scientists, and this contact between practical and skilled seamen and those versed in cosmography and astronomy was fruitful.

These sciences were also encouraged by the enlightened House of Aragon, under whose patronage Barcelona became a centre for the diffusion of Arabic knowledge, and therefore of advance in mathematics, astronomy and the construction of instruments.

This intellectual ferment was not without its influence on cartography, as may clearly be seen in that masterpiece, the Catalan atlas of about 1375.

Some attempts to extend the range of the portolan charts had already been made (for example, the chart of Angellino de Dalorto), and at about the same time Marino Sanudo was striving to reconcile the old and the new data.

The completion of this reformation of the world map was the work of the Catalan cartographers.

Precursors and origins

Though the Catalan atlas is the earliest complete example of its kind which has survived, it was undoubtedly preceded by others of similar general design.

The Medici sea atlas of 1351 contains a “world” map (extending eastwards as far as the west coast of India only) which resembles it in the outline of the coasts and in interior details.

From the nomenclature, it is probably of Ligurian origin. An even earlier chart (probably covering the whole “world” originally), that by Angelino Dulcert, of Majorca, dated 1339, also has points of resemblance to the Catalan atlas of 1375.

In view of the possible identity of Dulcert and Dalorto, and the Ligurian origin of the Medici atlas, we may conclude that this type of world map, though developed by Catalans, originated early in the fourteenth century in northern Italy, where the narrative of Marco Polo—which, as will be seen, supplied many of the details embodied in the map—would be most readily available.

The Catalan atlas (1375) and Cresques

We know in unusual detail the circumstances in which the Catalan atlas of 1375 (the date of the calendar which accompanies it) was produced and the career of the cartographer who compiled it.

When in 1381 the envoy of the French king asked King Peter of Aragon for a copy of the latest world map (proof in itself that the reputation of the Catalan school had then been widely recognized) he was given this example, which has been preserved at Paris ever since.

It is on record that it was the work of “Cresques le juif.” Abraham Cresques, a Jew of Palma in the island of Majorca, for many years was “master of mappae mundi and of compasses,” i.e. cartographer and instrument maker, to the King of Aragon, from whom he received special privileges and protection. There are several references to world maps executed by him, though this is the only one now known.

After his death in 1387 his work was carried on by his son, Jafuda, but the day of the Jewish school of cartography at Majorca was already drawing to a close, owing to the wave of persecution which swept through the Aragon kingdom in the closing years of the century.

Jafuda submitted to force and became a Christian in 1391, receiving the name Jaime Ribes, but his position was not improved thereby, and he left Palma for Barcelona.

There he continued his work, in increasingly difficult circumstances, until finally he accepted the invitation of Prince Henry of Portugal to take up his residence in that country, where he instructed the Portuguese in cosmography and the making of charts.

This link between the Majorcan school and the beginning of Portuguese overseas expansion is of obvious significance.

Patrons and sources

The patrons of Cresques, King Peter III of Aragon and his son, in addition to their scientific interests, were keenly interested in reports of Eastern lands in relation to their forward economic policy, and were at special pains to secure manuscript copies of Marco Polo’s “Description of the World”, the travels of Odoric of Pordenone, and, what may surprise the modern reader, the Voyage of Sir John Mandeville.

Though fabulous in part, Mandeville’s book has a scientific background. He was quite sound, for example, on the sphericity of the earth; as he says, “…who so would pursue them for to environ the earth who so had grace of God to hold the way, he might come right to the same countries that he were come of and come from and so go about the earth… few men assay to go so, and yet it might be done.”

The title of the atlas shows clearly the spirit in which it was executed and its content: “Mappamundi, that is to say, image of the world and of the regions which are on the earth and of the various kinds of peoples which inhabit it.”

The whole consists of twelve leaves mounted on boards to fold like a screen; four are occupied by cosmographical and navigational data, the remaining eight forming the map. Each leaf is 69 × 49 cm, so that the whole is approximately 69 cm × 3.9 m.

These proportions are of some significance, for they have undoubtedly restricted the cartographer in his portrayal of the extreme northern and southern regions.

This was perhaps to some extent deliberate—for two years before the composition of this map, we hear of the Infant John demanding a map “well executed and drawn with its East and West” and figuring “all that could be shown of the West and of the Strait (of Gibraltar) leading to the West.”

The Infant, in other words, was interested not in northern Europe and Asia or in southern Africa, but in the Orient and the western Ocean. The cartographer satisfied him by cutting out, as it were, an east–west rectangle from a circular world map which would cover the desired area.

Later Catalan maps, e.g. the Este map, retained the circular form.

The shape of the map, therefore, must not be taken as evidence on questions such as the extent or form of the African continent; nor does the change from a circular to a rectangular frame indicate specifically any change in ideas relating to the shape of the earth.

As an astronomer, Cresques accepted its sphericity.

Sources and influences

The sources of the Catalan atlas fall into three groups: (1) elements derived from the typical circular world-map of medieval times; (2) the outlines of the Black Sea, Mediterranean, and the coasts of western Europe based on the “normal” portolan chart; (3) details drawn from the narratives of the thirteenth- and fourteenth-century travellers in Asia, which transformed the cartographic representation of that continent.

The influence of the medieval world map may be seen in many features: Jerusalem, though not so strongly emphasized, is still approximately in the centre of the map; a portion of the original circumference of the circular map forms the coastline of north-east Asia, with the Caspian mountains still enclosing the tribes of Gog and Magog; the large island of Taprobane occupies approximately the same position as, for example, on the Hereford Map; the great west–east river beyond the Atlas Mountains resembles the traditional conception of the hydrography of North Africa, though contemporary names have been inserted. Clearly the contemporary additions are set in a much older framework.

Africa and traveller reports

The narratives of contemporary travellers were extensively used by the cartographer. The north-west coast of Africa extends beyond Cape Bojador to a point north of the Rio d’Oro.

An inscription records the departure of the Catalan Jacome Ferrer on a voyage to this “river of gold” in 1346, and some knowledge of the gold-producing region of the middle Niger is displayed.

The regional name Guinea (Ginuia), the Kingdom of Melli, and stages on the routes from Morocco to the Niger, e.g. Sigilmessa, Tebelt, T’agaza and Tenbuch (Timbuktu), are marked.

In north-east Africa, a knowledge of the Nile valley as far south as Dongala, where there was a Catholic mission early in the century, is apparent.

The delineation of the Nile system is vitiated, however, by the conception that it flowed from a great lake in the Guinea region. This lake may reflect rumours about the flood areas of the Niger, but the whole idea is very much older.

Asia: form and routes

It is, however, in its representation of Asia that the greatest interest of the Catalan map lies. For the first time in medieval cartography, the continent assumes a recognizable form, with one or two notable exceptions.

From the Caspian Sea in the west, with a fairly accurate outline in the style of the portolan charts, the Mongol domains stretch away eastwards to the coast of Cathay.

This makes a sweep from east to south with an approach to its actual form, and along it appear several of the great medieval ports and trading centres, frequented by Arab merchants.

In the interior are correctly placed the main divisions of the Mongol territory; from west to east, the “Empire of Sarra” (the Kipchak khanate), the “Empire of Medeia” (the Chagtai khanate of the middle), and the suzerain empire of the Great Khan, Catayo with its capital at Cambaluc (Peking).

If the map is stripped of its legends and drawings of the older tradition, it is apparent that the main interest of the compiler is concentrated in a central strip across the continent.

Herein lies a succession of physical features—mountain, river, and lake—and of towns with corrupt but recognizable forms of their medieval names as given in the narratives of the great travellers of the thirteenth century.

These are jumbled together in a manner sometimes difficult to understand, but with the help of Marco Polo’s narrative it is possible to disentangle the itineraries which the map was evidently intended to set out.

In the west is the Oxus river (fl. Organci) shown, as on most contemporary maps, flowing into the Caspian, and alongside it the early stages of the itinerary from Urganj (the medieval Khiva) through Bokhara and Samarcand to the sources of the river in the mountains of Amol, on the eastern limits of Persia.

These are the highlands of Badakshan where the route crossed the Pamirs. East of this lies the lake Issikol and Emalech, the seat of the Khan, the Armalec of other travellers, in the Kuldja region.

The delineation is then confused by the repetition of the Badakshan uplands, the mountains of “Baldassia,” a mistake which probably arose from a confusion over the river system of southern Asia.

Further east is “Chancio” (Kanchow, on the great loop of the Hwang-ho), and ultimately “Chambaleth,” the city of the Great Khan, and the goal of travellers from the west.

This, with several omissions, was in outline the route followed by Marco Polo’s father and uncle on their first journey to the Great Khan’s court.

It is also possible to discern traces of their second journey, accompanied by Marco, on a more southerly route through Eri (Herat), Badakshan, and along the southern edge of the Tarim basin from Khotan to the city of Lop.

The compiler, however, perhaps because he confused this desert area with the Gobi, has transferred this stretch to the north of the Issik Kul.

A third route is indicated rather confusedly on the extreme northern edge of the map. It is marked by a line of towns up the valley of the Volga from “Agitarchan” (Astrakhan) through “Sarra” (Sarai), “Borgar,” and thence eastwards through “Pascherit” (probably representing the territory of the Bash Kirds east of the middle Volga), and “Sebur,” or Sibir, a medieval settlement whose site is unknown but thought to be on the upper Irtish. In this quarter, the information on which the map was based was not drawn from Marco Polo.

To the south is a long east–west range, called the “mountains of Sebur,” representing the north-western face of the Tien Shan and Altai.

In the late thirteenth and early fourteenth centuries there were Franciscan mission stations at these localities, and the details no doubt came originally from the friars.

Cathay and ports

“Chambaleth,” the city of the Great Khan which so intrigued the chroniclers of the fourteenth century, receives due prominence, with a long legend describing its magnitude and grandeurs.

It stands near the apex of a triangle formed by two rivers and the ocean; each of the two rivers divides into three before reaching the sea, a representation embodying a somewhat confused notion of the interlinked natural and artificial waterways of China.

On the southern portion of the Cathay coast, the general uniformity of the coast is broken by three bays, and it is significant that these are associated with the three great ports: Zayton (near Changchow), Cansay (better known in medieval records as Quinsay, i.e. Hangchow) and Cincolam (Canton).

Of these, Canton is not mentioned by Marco Polo; it was, however, much frequented by Arab navigators and traders, upon whose reports the compiler was probably drawing.

The attempt at representing the configuration of the coast suggests at least that his informants were interested from a maritime point of view. Some of the islands off Quinsay may stand for the Chusan archipelago, and further to the south is the large island of Caynam, i.e. Hainan.

In the interior, the towns, in Cordier’s view, can in general be related to the itineraries described by Polo.

South-east of the coast of Cathay are numerous islands—“we are told that they number 7,548”—in which grow the spices. In the extreme corner is a portion of a great island which is named Taprobana.

A legend states that it is the last island in the East, and is called by the Tatars “Great Caulij.” Yule pointed out that Kao li was the name for Korea, and he therefore considered that the island depicted confused notions of the Korean peninsula and Japan.

India and the Indian Ocean

The delineation of the coastline of southern Asia has one major defect and one outstanding merit: the defect is the entire omission of the south-eastern peninsula; the merit is the portrayal for the first time of the Indian sub-continent in its peninsular form.

The first is difficult to explain; to make up for it the cartographer has inserted a great island of Java (mis-spelt Jana), which however was probably intended for Sumatra.

For the Indian peninsula, other sources are intermingled with Polo’s account. The Kingdoms of India as enumerated by Polo are absent from the map, and there are significant differences in the towns appearing in the two documents.

Conspicuous on the map is the “Christian kingdom” and city of “Columbo,” placed on the east coast. There is no doubt, however, that this is Quilon, on the west coast.

This form of the name (it is rendered Coilum by Polo), and other details, suggest that the compiler drew upon the writings of Friar Jordanus, who was a missionary in this area, and whose Book of Marvels was completed and in circulation by 1340.

In the area around the Gulf of Cambay, several towns are shown which are mentioned by Jordanus but not by Polo, e.g. Baroche and Gogo. There are other names, however, which are not found in Jordanus; but the commercial importance of Cambay (Canbetum), on the map, would account for the relatively detailed information about this region.

There is, however, no indication of the great river Indus, a striking omission also from Polo’s narrative. This probably arose from confusion between the Indus and the Ganges.

For the portion of the Indian Ocean included in the map, sources other than those embodied in Polo have been used.

The Persian Gulf, extending almost due west, has an outline similar to that on the Dulcert map, but is otherwise superior to any earlier map. In the Gulf, the “island of Ormis” (Hormuz) is shown, opposite the former settlement of the same name on the mainland.

The southern Arabian coast has names differing from those given by Polo, and in one of them “Adramant” we may recognize the modern Hadhramaut.

The island of “Scotra,” an important stage on the trade route from Aden to India, is misplaced to the east, and appears to occupy the approximate position of the Kuria Muria islands.

For India and the ocean to the west, therefore, we may conclude that charts were used which differed in detail from Polo’s account, though similar in general features.

That such charts existed we know from Polo’s own statements. Possibly additions were also made so that the map might serve as an illustration to his narrative.

The Este map and later Catalan maps

The only complete Catalan world map other than that of 1375 which has survived is the Este map preserved at Modena.

This map is circular, and although almost a hundred years later, it is clearly related to the Atlas of 1375. This resemblance in the content of the two maps strengthens the contention that the latter was derived from a circular prototype.

The nomenclature and the numerous legends, mostly in Catalan with a few in corrupt Latin, are often very similar to those of the 1375 atlas.

In some instances the legends are more complete, in others they are less detailed; they suggest therefore not direct copying but a common source.

This similarity is also evident in the delineation of the main features—most of those in the 1375 atlas are to be found on the Este map.

The northern portions of Asia and Europe, which lay outside the limits of the Catalan atlas, significantly contain very little detail.

On the southern coastline of Asia there are some differences, generally slight, between the two maps. The peninsula of India is much less pronounced on the Este map, and to the south is the large island of “Salam” or “Silan” (Ceylon) which fell outside the limits of the Catalan atlas.

A legend refers to its wealth in rubies and other precious stones. There can be no doubt, however, that the two outlines are fundamentally identical.

To the east is the island of “Java,” as on the Catalan atlas. The island of “Irapobana” is much enlarged, and is placed on the south-eastern margin of the map.

The surrounding ocean, the “Mar deles indies,” is filled with numerous nameless and featureless islands.

Africa in the Este map

Africa occupies most of the southern half of the map. The continent ends in a great arc, conforming to the circular frame of the map, and extending eastwards to form the southern boundary of the Indian Ocean.

On the west, a long narrow gulf from the circumfluent ocean almost severs this southerly projection from northern Africa. The southern interior is blank save for the legend: “Africa begins at the river Nile in Egypt and ends at Gutzola in the west: it includes the whole land of Barbaria, and the land in the south.”

This outline and legend have been interpreted to imply some knowledge of the southern extremity of Africa, and perhaps of a practicable route from the west to the Indian Ocean.

That the great western gulf reflects some knowledge of the Gulf of Guinea is more probable.

The design of the northern half of the continent in general resembles that of the other Catalan charts, but the north-western coast embodies some details of contemporary Portuguese voyages as far as “C. tide” (Cape Verde) and “C. groso.”

From this evidence, the map is usually dated about 1450. Near the gulf are the Mountains of the Moon, from which five rivers flow northwards to a lake on the “western Nile.”

This lake probably represents the area around the Upper Niger liable to inundation; Dr. Kimble has pointed out that these rivers may well represent the five main sources of the Niger.

These Mountains of the Moon are stated to be on the Equator, and the streams are called the “riu de lor.”

We may therefore assume that the headwaters of the Niger marked the approximate limit of knowledge in this region, and it is not improbable that reports of the sea to the south had been received.

These may have induced the cartographer to accept the western gulf of Ptolemy, but to enlarge it considerably. The name “river of gold” recalls the inscription on the Catalan Atlas.

The portrayal of the interior thus goes back at least to 1375. Apart therefore from a small portion of the coastline, the map owes nothing to Portuguese exploration.

Conservatism and critical realism

Some surprise has been expressed that a map of 1450 should contain relatively up-to-date details with antiquated ideas in other areas, and this has produced some rather involved explanations.

Taking into consideration the lack of details and names in the southern regions of Africa, we may plausibly conjecture that, as an exception to the usual conservatism, the draughtsman, in Africa at least, had removed all the detail for which he had no evidence, to obtain a framework on which to insert the latest Portuguese discoveries.

It must remain debatable whether the outline of the southern extremity represents some knowledge of the Cape. The outline may be entirely imposed by the frame of the map: at the most, it may reflect the kind of report that we find on Fra Mauro’s map.

The merit of the Catalan cartographers lay in the skill with which they employed the best contemporary sources to modify the traditional world picture, never proceeding further than the evidence warranted.

In the same spirit they removed from the map most of the traditional fables which had been accepted for centuries, and preferred, for example, to omit the northern and southern regions entirely, or to leave southern Africa a blank rather than to fill it with the anthropagi and other monsters which adorn the medieval maps.

Though drawings of men and animals still figure on their works they are in the main those for which there was some contemporary, or nearly contemporary, warrant; for example, Mansa Musa, the lord of Guinea, whose pilgrimage to Mecca created a sensation in 1324, or Olub bein, the ruler of the Tatars.

In this spirit of critical realism, the Catalan cartographers of the fourteenth century threw off the bonds of tradition, and anticipated the achievements of the Renaissance.