Mathematical Geography and Cartography in Islam and Their Continuation in the Occident I, Historical Presentation Part I

by Professor Fuat Sezgin

Being an English Version of Volume X of Geschichte Des Arabischen Schrifttums, Translated by Guy Moore and Geoff Sammon, 2005, Institute for the History of Arabic-Islamic Science at the Johann Wolfgang Goethe University, Frankfurt am Main.

This article is extracted from Fuat Sezgin’s book: Jubilaumsband zum dreiBigjahrigen Bestehen des Institutes fur Geschichte der Arabisch-Islamischen Wissenschaften In europaischen Sprachen erschienene Vorworte zu Publikationen des Institutes aus den Jahren 1984 bis 2011, Institut fur Geschichte der Arabisch-Islamischen Wissenschaften an der Johann Wolfgang Goethe-Universitat, Frankfurt am Main, 2011, pp. 63-82.

Some readers of the Geschichte des arabischen Schrifttums may initially experience a feeling of disappointment. First of all, after waiting for so long for the volumes at hand to be published, they find those achievements made by the Arab-Islamic cultural sphere in the fields of anthropogeography which have been adequately discussed by Arabist research to be missing from this work, and, secondly, most of the topics contained in the table of contents do not immediately seem to be in line with those treated in the previous volumes of my work. I do hope, however, that in the course of following my explanations and lines of reasoning, readers will both understand why I chose this approach as well as why I inevitably had to address this subject matter in a much more extensive manner than had been the case with topics tackled in previous volumes.

Professor Fuat Sezgin

Professor Fuat Sezgin

The history of mathematical geography and cartography is particularly suitable for delivering a convincing conception of the unity of the history of science. The idea of a principal stream running through the history of the development of the sciences which I have gained in the course of dealing with the history of the mathematical and scientific disciplines in the Arab-Islamic cultural world has been substantially intensified and consolidated while working on the volumes at hand. In the face of such conviction, there neither remains room for heuristic considerations nor for considering the contributions made by a cultural sphere which has long of necessity taken a receptive or imitative stance in the history of the sciences to be inferior or worthy of contempt. Viewed from a perspective shaped by this conviction, the Greeks are seen to hold an outstanding position in the history of the sciences which meets with the respect and admiration of historians. Some centuries after appearing on the historical stage, the intellectual history of humankind bestowed upon them, in a specific era and under favourable conditions, the role of further advancing the knowledge which had been gained in previous civilisations, and of codifying, defining and extending this knowledge as independent disciplines. Indeed, they were able to do this several centuries long with astonishing intensity and creativity, until their efforts came to a standstill for reasons to be found in their history.

After a somewhat slow reception of those sciences which, not without some contributions of their own, had been mainly pursued by the Byzantines, Syrians, Persians, and Indians, it was the Arabs – or the Muslims to be more precise – along with non-Muslims living and working under their rule who appeared as early as in the 7th century, still within the context of this process of reception, and demonstrated that they, albeit under new historical, geographic and economic conditions, were the true pupils of the Greeks. Within a relatively short period of time, most of the surviving Greek works had been translated into Arabic. The rapid reception and assimilation of the sciences adopted from the Greeks and from other civilisations had to overcome no obstacles of political or religious nature. For example, the adoption of the Greek conception of a spherical Earth occurred as early as at the beginning of the 8th century, without any religious opposition. It was without reservation that Aristotle was called the “first master” (al-muᶜallim al-auwal). Therefore, it should come as no surprise that even before or by the middle of the 9th century at the latest, the Islamic world was already creatively active in almost all fields of science. This activity then continued over centuries, in some disciplines probably proceeding through into the 16th century.

As early as during their prime in the 10th century, these sciences began their journey, in the form of Hebrew and Latin translations, from Arab Spain to non-Spanish Europe, while in the form of Byzantine-Greek translations they took their route to Byzantium. As early as in the 11th century, many Arab works could be found there in Greek versions and under the names of Greek authorities. In Western Europe, the process of translating these works increased in intensity and extent in the 12th century and led to a long-lasting assimilation of Arabic works in the 13th, 14th and even in the 15th century. It was only in the 16th century that any European-own creative activity could be perceived. Given such a notion, the conception of the Renaissance as a “return of classical antiquity” in the field of fine art may be justified; however, where this applies to the history of science, I hold that this view is fundamentally incorrect and doomed.

Arabists have already provided a number of clues for consideration of the question of whether one could not for once also treat the problem faced by the history of geography regarding the creation of a perfect pan] cartographic depiction of the Mediterranean and the Black Sea, a question which has long resisted all answers, i.e. the question as to the unexpected appearance of the so-called portolan charts (in around 1300) and the cartographic development which followed in their wake in Europe, from a perspective such as this. However, these clues failed to induce a change of opinion among cartography historians with a Eurocentric attitude. In general, these clues stem from the early phases of Arabist research and within the context of the non-Arabist treatment of the history of geography remain either unknown, apart from some few exceptions, or have been ignored or not taken seriously. Yet, it must be said that pioneering works in this field continued to lack the support of Arab-Islamic maps. In addition to this, it took a long time before the work of al-Bīrūnī (died 1048) became known, in which he established mathematical geography as a discipline of its own. In fact, it took until 1939 before this work was discovered and it was only made available to research in the form of an edition and English translation in the second half of the 20th century.

In preparing the volume of my Geschichte des arabischen Schrifttums dedicated to Arab geography, I was guided by the commendable work of my Arabist predecessors, father and son Jean-Jacques, respectively, Louis-Amélie Sédillot, Joseph-Toussaint Reinaud, Joachim Lelewel, Carlo Alfonso Nallino and Carl Schoy to the field of mathematical geography, and thus to exploring the history of this discipline as practised in the Arab-Islamic world, to its history, in general, and its continuation in the Occident, to the question concerning its expression in maps produced within its own civilisation and the traces which it may have left in European maps. I allowed myself to be led deeper into this field and developed a fascination. At the present point in time, now that some fifteen years have passed, I must content myself with what has been achieved, since it has become clear to me that I have only been able to advance the work which needs to be done in this field by a single step. In anticipation of what is contained in these two volumes, I can describe the results of my findings as follows:

  1. The field of mathematical geography, first pursued by the Greeks and then developed into an independent discipline in the Arab-Islamic culture area in the first half of the 11th century, experienced here an uninterrupted process of development from the mid 8th century through into the 16th century. The degrees of longitude and latitude produced in the course of this process have continuously found expression in maps. It is, in my opinion, not possible to write an accurate history of this discipline without taking note of the accomplishments of the Arab-Islamic culture or of the material which Arabists have already made available. The volumes at hand represent a first such attempt.
  2. There are a number of problems faced by European geography and discussed over the past one and a half centuries or so which cannot be explained from the perspective of the European tradition. Rather, these can only be solved in the light of the mathematical geography and cartography intensively practised since the 9th century in the adjacent Islamic culture. This includes the creation or origin of the so-called portolan charts and their linear networks, or the provenance of the reasonable and ever-improving depictions of Africa in European maps which began to appear as from the 13th century, or the accurate cartographic depiction of the triangular Indian subcontinent, or the initially “dragon tail-shaped” and later virtually correct depiction of South-East Asia, or even the mapping of Madagascar. In short, we are talking about the depiction of the whole Indian Ocean, plus the configuration of Asia and its constituent parts, such as Russia, Siberia, the Caspian Sea, and the other Asian inland waters with their oro- hydrographical systems, the depiction of Arabia, the Red Sea, Persia, the Persian Gulf, Anatolia, the Black Sea, and the Balkan Peninsula.
  3. Many of the maps stemming from the various stages in the development of Arab-Islamic cartography, of which most have been lost except some few surviving examples, can be restored from preserved European maps originally produced by copying and to a certain extent modifying models from the Arab-Islamic world. For example, the two maps of Northern Asia published in Holland in 1726 together with the French translation of the book on the genealogy of the Tartars written by the statesman Abu l-Ġāzī Bahādur Ḫān (died 1663) reproduce translations of local original maps made on the initiative of Philipp Johann von Strahlenberg. One of these maps presumably dated from the 13th century, the other from around 1550 (v. Map Nos. 107, 130). The two perfect depictions of Africa which bear the names of Alberto Cantino (v. Map No. 191) and Nicolo di Canerio (v. Map No. 192) are in all probability nothing other than Portuguese or Italian versions of a model stemming from the Islamic world which had either found its way to Portugal before the expedition by Vasco da Gama or had been brought back by him from that expedition. The fact that the “Javanese” maps which fell into the hands of Portuguese mariners, as these themselves claimed, during the conquest of Malacca were immediately translated into Portuguese and sent to King Joao II in Lisbon as the most beautiful maps that had ever been seen means that we can without reservation- albeit against the background of the newly-gained conception of the cartographic-historical development of the whole region – see in this one the most recent depictions produced, shortly before the Portuguese conquest of the Malay peninsula, as part of the Arab-Islamic cartography in the southern hemisphere of our Earth. I do not know when the practically perfect shapes of Madagascar and of the South-East Asian archipelago which appear on these maps were further improved. In the 19th century, or perhaps even only as late as in the 20th century?

The form given to the Caspian Sea on European world maps since the 14th century is sometimes good, sometimes poor, and sometimes it is even depicted twice on one and the same map; by contrast, however, it is perfectly depicted on regional maps. This depiction then, with some few exceptions, had to make way over a period of some 200 years to a melon-shaped version which extended around 17° too far in east-west length. This version, based on the Ptolemaic Geography, continued to circulate until, towards the end of the 17th and the first quarter of the 18th century, the “realistic” depiction was allowed to reappear, in the same manner as it had done on the 300-year-old regional maps. Is there not some justification for assuming that these regional maps and their accurate depiction of that body of water actually represent a shape mathematically recorded in the Arab-Islamic civilisation, if irrefutable arguments serve to support this assumption?

In this respect, mention should also be made of a map of Persia drawn up in 1724 by Guillaume Delisle. If this French court cartographer working from his Parisian studio managed to draw up a map of Persia with an unparalleled degree of precision for the coastal lines and mainland outlines and with the geographical positions of some 600 places, which today prove to be largely accurate and even very precise, if this map reveals clear traces of transcription and translation, and if its coordinates prove to coincide with data contained in the most recent Persian geographical tables of the time, where these have survived, should we not then think of it being the translation of just one of these most recent Persian maps? Otherwise, we would have to ask ourselves how else a cartographer could have managed such an accurate and detailed mathematical recording of a large, far-distant part of the Earth.

In order to make it easier for readers to follow this argument, which could perhaps be understood to be aphoristic, here some essential aspects regarding the process of development of mathematical geography and cartography in the Arab-Islamic culture. The world map created by the geographers and astronomers in the service of Caliph Maᵓmūn in the first third of the 9th century and rediscovered about 15 years ago provides new incentive for the treatment of the history of these disciplines in Islam and, over and above this, delivers a certain base and suitable starting point for present-day research. The scholars commissioned by the Caliph were responsible for drawing up a new world map and of writing a geographical work to accompany it on the basis of the Ptolemaic Geography and their own measurements and observations. In all probability, the Ptolemaic Geography neither originally nor at the time included maps. On the other hand, it would seem that those scholars did have the world map of Marinos at their disposal, which had served as a basis for the Ptolemaic conception of the Earth. Knowledge gained in the time after Ptolemy in late antiquity, the measurements and observations carried out by the Maᵓmūn geographers themselves, and finally the future collaboration between the many scholars which Hipparchus had dreamt of, all paved the way for a somewhat modified, or rather, a much better depiction of the shape of the Earth. The original world map produced by the Maᵓmūn geographers was in all probability based on stereographic projection. The version surviving from 1340, which certainly resulted from multiple copying of the original, bears a globular projection (v. Map No. la). A similar projection only appeared on European world maps as from the first half of the 16th century. The map by the Maᵓmūn geographers (v. Map No. lb), which can be fully reconstructed on the basis of the data for 3000 or so localities taken from its table of coordinates, which also survived, offers research into the history of cartography a sound basis for a future, more detailed evaluation of how it continued to develop in Islamic culture and beyond.

The world map and its table of coordinates found their way to Europe via Spain in the first half of the 12th century at the latest. The oldest indirect European imitation of the world map is contained in the encyclopaedic work of Brunetto Latini (around 1265, v. Map No. 55). Its copyist painted the map depiction, but obviously failed to concern himself with the place names written in an alien language. I believe, I have also found an interesting imitation of the table of coordinates from the first half of the 15th century produced in the German-Austrian language area. The length of the Mediterranean on the world map, which had already been reduced by 10° to approx. 53° in comparison with Ptolemy (although still too long), and some further unrealistic dimensions, which for their part were already considered outdated in the Islamic world, continue to appear on European world maps into the 16th century, and together with Northern Asia even into the first quarter of the 18th century.

Some historians of cartography who brought the emergence of the portolan charts and, especially, the world map by Marino Sanuto -Petrus Vesconte (around 1320, v. Map No. 56) into connection with the round world map from the geographic work of al-Idrīsī (1154, v. Map Nos. 4, 5, 6), which the latter, who came from Ceuta, had created on behalf of his host, the Norman King Roger II of Sicily, inevitably came to the assumption that al-Idrīsī had made use of the “Ptolemaic” world map as a basis for his work. In explaining the shape of the southern coasts of Asia, the Red Sea, or the Arabian peninsula, and of the Persian Gulf on “Portuguese” maps of the Indian Ocean, some researchers proceed in similar fashion, when they directly bring these maps into connection with the “Ptolemaic” world map. The recently-discovered Maᵓmūn map provides a plausible explanation in all these cases, as well as in particular in the case of the large peninsula of South-East Asia called the “Dragon’s Tail”, which could be found on European and “Portuguese” world maps for more than two hundred years.

Not only the Maᵓmūn map and its table of coordinates found their way to Europe at a relatively early stage. In the first half of the 12th century, further, newly acquired coordinates reached non-Spanish Europe together with processes for determining degrees of longitude and latitude which had already been known to the Greeks as well as some methods developed by Arab-Islamic astronomers themselves for determining degrees of latitude and the altitude of the pole. All this stood in Hebrew and Latin translations of Arabic handbooks on astronomy. It should also be noted that the results of observations carried out between Baghdād and Toledo, which led to the length of the Mediterranean being reduced by 18°30′ to approx. 44°30′ in comparison with the Ptolemaic Geography and the prime meridian being shifted from Toledo out into the Atlantic by 28°30′ had equally been known in Europe since the 12th century. Since the time when the westward shifted prime meridian (occidens verum) had been treated by Roger Bacon in his Opus maius, the knowledge of that meridian is to be found in numerous geographical tables which, for their part, represent heterogeneous registers of data compiled from Arab models of varying origin.

In the course of time, not only the number of coordinate tables increased in Europe, but so too did the ability to determine degrees of longitude and latitude. With regard to the old world, the coordinates determined in Europe were mainly restricted to Western, Southern and Northern Europe up until the end of the 17th century. In some cases, they were of a high quality, although they continued to be listed together with coordinates from Arab sources of varying provenance, yet still without taking the differing prime meridians on which they had been based into account.

It should be noted in this respect that the tables of coordinates compiled in Europe were not used for the purpose of drawing up maps, with the exception of scholars in Spain who stood in the tradition of the Arab-Islamic geography. Perhaps Johannes Kepler was the first European scholar to attempt to produce a correct depiction of the Mediterranean as part of a world map with the help of the coordinates available to him (v. Map No. 99). His attempt was doomed to failure from the outset, since such a task based on a major geographical table of heterogeneous coordinates was impossible to accomplish.

It would seem that Wilhelm Schickard (1592-1635) was the first geographer and astronomer to unequivocally voice his misgivings about those geographical tables and the shortcomings of the maps of the day. However, the project he started with a view to producing a precise map of the old known world on the basis of the comparative geographical tables of Abū l-Fidāᵓ and consultation of further Arab geographical works shows that he, as all his European predecessors had been, was unaware that the book by Abū l-Fidāᵓ merely provided the status of the mathematical recording of the Earth’s surface in the tradition of the Maᵓmūn geographers up until around 1250. Abū l-Fidāᵓ had not yet been aware of the geographical tables which [XIX] recorded the important corrections made between Toledo and Baghdād since the first half of the 11th century. Unfortunately, Schickard, in much the same way as all his predecessors and successors, did not know either that astronomers and geographers working in around 1270 in Marāgha, the seat of the government of the western Mongolian dynasty, had come upon the idea of integrating the degrees of longitude between Toledo and Baghdād, which had been corrected and whose number had been continually increased by the western school of Islamic geography, with those which had been counted eastwards from Baghdād. The decisive requirement for this correction had also been created in the first half of the 11th century, namely by al-Bīrūnī. In his unique undertaking, which involved him personally measuring the distances between Ghazna and Baghdād across two routes (2 times approx. 2000 km), he used astronomical observation and the rules of spherical trigonometry to determine with astonishing precision the coordinates of some 60 localities. The errors in the longitudes he measured range from just 6 to just 40 minutes. The graticule of the eastern part of the Islamic world was continually advanced and extended by following generations. The two most significant geographical tables thus produced with integrated degrees of longitude, namely those by Naṣīraddīn aṭ-Ṭusī (died 1274) and Uluġ Beg (died 1449), were only made accessible in Europe in 1652 by Johannes Gravius.

The fact that such a graticule not only appears on the surviving Arab-Islamic graduated maps of Asia from the period between the 14th and 18th centuries, but also on numerous European maps of Asia and even the world unfortunately seems to have escaped the attention of historians of cartography. And even though the assumption that the highly-praised, ungraduated map by Fra Mauro of 1459 (v. Map No. 63) had itself already been based on a model equipped with such a graticule (with a far-reduced east-west extension of the known world) must remain mere conjecture, it can be proven, through the use of Arab-Islamic models for the creation of the maps of Asia and Africa by Gian Battista Ramusio (1550-1559, v. Map Nos. 205a-c) and the world and Asian maps by Giacomo Gastaldi of 1559-1561 (v. Map Nos. 96b, 113a-d) that a graticule of this kind had found its way into maps of Asia and the world later produced in Europe. It should not surprise us all too much that this graticule was still managing to assert itself on many European maps of Persia and Asia in the first quarter of the 18th century, although in the Paris of the end of the 17th century, a new prime meridian had been integrated which passed through the French capital. Neither should it surprise us when the graduated maps of Siberia known to us from the last quarter of the 17th century and the first quarter of the 18th century partly lead to traces of this graticule and partly to the world map of the Maᵓmūn geographers.

Historians of geography have so far practically disregarded this factor altogether. They have simply not asked the question as to the origin and significance of the graticules which appeared on European world and regional maps of the old known world as from the beginning of the 16th century. Perhaps the one or the other of them has contemplated this question, but a satisfactory explanation would not have been easy to find, as long as the graticules could not be brought into connection with the practice of cartography in the Arab-Islamic culture. In this respect, I find it would seem to be of great interest to discuss a special kind of graticule which can be found on a number of European maps, which prove to be direct translations of models from the Arab-Islamic world, and which we also encounter on some surviving maps of that culture. On these maps, the graduation is produced by entering the figures of the degrees of longitude and latitude to scale at the edge of the map. The distances between these numbered lines at the northern and the southern edge diverge to such an extent that they equal the appropriate meridians, without the meridian lines having been drawn as solid lines. The marks for the degrees of latitude at the eastern and western edge are equidistant, whereby the lines of the degrees of latitude are also not solid lines. All in all, the degrees of longitude and latitude marked at the edge represent a globular projection. For example, this kind of graduation appears on the maps of Persia by Gastaldi (v. Map Nos. 113b, 113c), on the map of Asia and the map of Persia by Nicolas Sanson (Map Nos. 118b, 138b) and on the Ottoman maps of Egypt, Anatolia and the Black Sea (Nos. 45, 46b, 155a).

In order to round off this general insight into the ideas discussed in the two volumes at hand, which this preface sought to provide, the following pages present some details on the stages along which the Arab-Islamic mathematical geography and cartography were continued in the Occident. Looking back at the first half [XXI] of the 16th century, that time in which Asia was depicted in a completely new way on the maps created by the eminent geographers Giacomo Gastaldi, Abraham Ortelius and Gerard Mercator, I will remind my readers of the remark made by Ortelius to the effect that Gastaldi had produced his depiction in the tradition of Abū l-Fidāᵓ. The fact that Ortelius believed that he had no choice but to refer to the tradition of Abū l-Fidāᵓ in order to be able to explain the origination of Gastaldi’s map of Asia (1561), whose new graticule and wealth of topography and toponymy led to the break with the dominance of maps bearing the name of Ptolemy which had prevailed since 1477, is highly significant. We know today that the comparative tables of coordinates produced by Abū l-Fidāᵓ would not have been sufficient or suitable as a basis for the Gastaldi map. Rather, this map must have reproduced in a somewhat distorted manner an Arab-Islamic cartographic model, albeit together with the melon-shaped depiction of the Caspian Sea in the manner of the Ptolemaic Geography. Ortelius and Mercator for then part used the Gastaldi map of Asia for their own Asian maps, with certain modifications and also additions, namely for Siberia. It is typical of the way in which these cartographers worked that, for example, the place which Arab geographers and historians called Sibir, where later, in 1576, the city of Tobolsk was established, appeared on an Arab-Islamic map originating in all probability from the 13th century and on an Ottoman geographical table produced in 1524 as having the coordinates of Long. 87°30′, Lat. 58°30′, and then appeared on the map by Ortelius under the name of Sibier at approx. Long. 97°, Lat. 68° and on the Mercator map at approx. Long. 97°, Lat. 58°. Or, to give another example, they position the city of Hanbaliq (Cumbalich, Peking) and the “Chinese Sea” Kitaia Lacus a couple of thousand kilometres to the west, now placed north of the Caspian Sea. From this one may conclude that those great cartographers were, on the one hand, working with models of Arab-Islamic origin, while, on the other, when drawing up their own maps they were not immune to wrongly placing towns or other cartographic data or of adopting errors made earlier.

So, what did the routes look like by which depictions produced by Arab-Islamic cartographers found their way to Europe? The earliest adoption of world maps, such as that by Brunetto Latini (around 1265, Map No. 55), or, shortly afterwards, the maps of the Mediterranean and of the Black Sea, which were later described as portolan charts, probably came to Europe after Italians and the French had visited Arab Spain or as a result of encounters between Europeans and Arabs. Maps only sporadically and randomly found their way to other parts of Europe. This also applies to those maps which were purchased from encounters between inhabitants of the Eastern and Southern Mediterranean coasts and merchants and traders on their travels or to maps which had been looted during the Crusades. Clear and informative details are given by Portuguese sources and sailors from the first four decades of the 16th century, both from the Portugal before the expedition undertaken by Bartolomeu Dias, as well as from the time thereafter, about maps and the related aspects of navigating in the Indian Ocean. This includes reports on maps which had been acquired from the Arabs and which were forwarded by courier to the Crown, remarks to the effect that these maps were fairly often graduated, and served mariners in conjunction with compasses which were equipped with the suspended device later called the Cardan suspension, and, finally, remarks to the effect that the sea route around the tip of Southern Africa, the Cape of Good Hope, and the island of Madagascar, had long been known. It was an unfortunate stroke of fate in the study of the history of geography that led to a trend developing in the last two centuries in which the perfect maps of Eastern Africa, in particular, and of Africa, in general, which had come into circulation in Portuguese, Spanish and Italian versions shortly after the first return of Vasco da Gama, are viewed as maps based on the knowledge gathered, observations made and measurements taken by Vasco da Gama, his crew, and other mariners. Such a view is unrealistic and no longer tenable today.

The process of the acquisition of depictions and representations from the Arab-Islamic cartography took on a new character in the Europe of the 17th century and gained in intensity. The increasing contacts between European travellers of the 17th century with the cultural centres of the Islamic world in Northern Africa, in the Ottoman Empire, in Persia, India, Central Asia, and Russia were decisive for this development. Let us call to mind that, as late as in 1635, Wilhelm Schickard did not even know that maps existed in the Islamic world and that the most up-to-date of them would have served him better in examining the correctness of European maps, which he viewed with great scepticism, and in pursing his goal of drawing up a more accurate map of the old known world than the book by Abū l-Fidāᵓ which he tried to borrow for years and which largely contained coordinates already outdated in the Islamic world.

In 1636, one year after Schickard’s death, Adam Olearius (1599-1671) from Gottorp/Schleswig travelled as a member of a trade delegation via Moscow to Persia. During his stay in Shemakha, the capital of Shirvan (30.12.1636-1637), a muezzin from Hijaz showed him a table containing degrees of longitude and latitude of almost all of Asia as well as a number of regional maps. He merged the maps of Persia and Anatolia, transcribed them into Roman characters and added them to the edition of his travel account (1646). Although his colleagues at Leipzig University and his friends took it amiss that in his new depiction of Persia, and especially of the Caspian Sea, “he departed from the famous old geographers Ptolemy, Strabo, Dionysius Alexandrinus and others”, this oldest map, definitely transcribed from the Arab-Islamic cultural sphere, nevertheless soon took effect in Europe. It is very astonishing that the historiography of geography has not yet taken note of this work as the example of a transcription of an Arab-Islamic model and that the appropriate conclusions have not been drawn from this. Whatever the case may be, the way Olearius proceeded seems to have fallen on fruitful soil with his contemporary, the French court geographer Nicolas Sanson. Sanson’s general and regional maps of Asia change with the publication of his map of “Tartaria” (1654). In the legend to this map, he notes that he had also made use of Arab maps as models. Sanson does indeed use a graticule whose prime meridian, as in the map translated by Olearius, lies 28° or 28°30′ west of Toledo. His 1659 map of Siberia marks the first time that an accurate depiction of the Caspian Sea was again shown after this depiction had fallen into oblivion due to the spread of maps bearing the name of Ptolemy. One can well imagine that in a world city such as Paris, where good political and economic relations were maintained with the Ottoman Empire, with the Safavids in Persia and with the Mogul Empire in India, specific endeavours were undertaken to obtain maps from those countries. Such actions were supported by scholars travelling to these countries for stays of several years’ duration, including Orientalists. Of the young French scholars who brought back maps, geographical tables and books to Paris in the second half of the 17th century and the first quarter of the 18th century, mention must be made of Jean Chardin, Melchisédec Thévenot, Jean-Baptiste Tavernier, Francois Pétis de la Croix and his like-named son, Francois Bernier and Jean-Baptiste Fabre. The maps which these and other travellers brought back with them either came into circulation, once they had been translated, or, in the case of regional maps, were worked into general maps. For example, Jean-Baptiste Fabre (died 1702) brought back a map of the Black Sea from Constantinople which enjoyed great fame on the Bosporus and which was also admired in Paris, both as a separate map and when integrated into general maps. We can well imagine that this was not a one-off case. When the Orientalist Adrian Reland, a contemporary of the above-named scholars, says about his map of Persia (v. Map No. 139a), which he published in 1705 and which included a remarkably good depiction of the Caspian Sea, that he had created it on the basis of the works of the “greatest Arab and Persian geographers”, then he must certainly mean that he had produced his map as a reproduction of those maps contained in the works of those geographers. Based on this map and on the models translated or brought back by Olearius and Tavernier, Johann Baptist Homann, as he said himself, produced his famous map of Persia. Only rarely do we learn anything about the direct models used for the cartographic compilations or treatments produced in Europe. In this respect, the English mapmaker Emmanuel Bowen with his detailed legend to his 1738 map of Anatolia, the Black Sea and Persia proves an exception to the rule.

In this context, it is informative to read how in 1755 the historian of geography and cartographer Robert de Vaugondy complained about the actions of his contemporary Jean-Baptiste d’Anville in his depiction of Anatolia and Persia:

“We would gladly know the originals which provide the new materials on these countries given by Mr. d’Anville in the first part of his [map of] Asia. These contain details which are different to those which one may expect from a travel account. The topography which they offer can only have been taken from regional maps which were drawn up locally and where knowledge of these would undoubtedly be very useful to us.”

The enlightenment which Vaugondy missed as to the sources was later impressively added by d’Anville in his two works on the geography and cartography of India (1753) and of Egypt (1766). From this, we can see that Arab, Persian and Turkish historical and geographical works, tables and maps formed his major sources. It is merely regrettable that d’Anville failed to recognise the true reason for the seemingly over-sized degrees of longitude in the maps and tables from the Arab-Islamic world and therefore only took their degrees of latitude into consideration. His successor, James Rennell, was able to draw upon a much broader base of sources from the Islamic culture for the treatment of his geographical writings and maps on India. Although he too failed to uncover the secret of the seemingly overlarge degrees of longitude, he did find a clever solution through his trust in the coordinates entered into the Arab-Persian tables. He avoided the problem of the overlarge degrees of longitude, which had been created by shifting the prime meridian westwards by 28°30′ from Toledo or 17°30′ westwards from the Canary Islands, by counting the localities in Persia and India of interest to him from Aleppo. It was as if Rennell had, as it were, made the meridian which passes through Aleppo into his prime meridian. In comparing the results he had gained in this manner with some of the few degrees of longitude which had, up until then, been determined by observing the moons of Jupiter by telescope, he discovered that they revealed an astonishingly high degree of accuracy. He formulated his admiration in the words: “Had Ptolemy lived in the present times, he might have expressed his wonder, that, considering the advantages we possess, our maps of this part should be so incorrect; when the tables of Abulfeda, Naṣīraddīn, and Uluġ Beig, and the History of Timur, by Sherefeddīn, have been so long amongst us, in an European language” (see p. 610 below).

Rennell, d’Anville and their predecessors Strahlenberg and Delisle also wrote books to accompany their maps in which they endeavoured to describe their methods and practice and, more or less, the topography of the mapped areas. Over and above this, the works by the first two also bear a humanist character in their view of geographical history. They give us an astonishingly clear and unfalsified picture of the objectives and limits of the contributions which an individual scholar could perform in the 18th century. We are informed of how they endeavoured to verify and supplement the general maps contained in European editions, where the models created in the Arab-Islamic culture were either no longer obtainable or had been lost, by studying surviving, local Persian, Arab or Turkish regional maps (or their translations, imitations, or adaptations), tables of coordinates and translations of geographical or historical works. Besides these, they also made use of the few degrees of longitude which had been determined by observation of the moons of Jupiter and of the degrees of latitude which a number of Jesuit priests had found out. To illustrate their use of regional maps, which were considered trustworthy, we mention the Ottoman depiction of the northern half of the Red Sea (drawn around 1538) and that of Lake Urmia, which d’Anville adopted. The first shows the Gulf of cAqaba, erroneously divided into two (v. Map No. 208), a depiction apparently taken from the Turkish model. The second, incorporated into d’Anville’s general maps, shows that the original Turkish expressions of “Donkey’s Island”, “Sheep’s Island” and “Goat’s Island” in Lake Urmia had been left standing in European transcription.

This is the way in which from the beginning of the 18th century onwards Europe’s emerging creative contribution to the cartographic depiction of Asia and Africa could be seen to be setting in, and whose most recent stage we now know. This process was not a quick one, did not always proceed in linear fashion nor without setbacks and caesuras. The great Sicilian Orientalist Michele Amari bemoaned as late as in the second half of the 19th century that no map of his home island had yet been produced by triangulation. He had to balance the lack by using materials of Ptolemy and the Arabs which were accessible to him (v. vol. II, pp. 35-36).

The extraordinarily useful information and evidence given by d’An-ville and Rennell and the brilliant geographical-historical views expressed by the astute Alexander von Humboldt on the role of the Arab-Islamic geographers which he had formed by reading their works in translations and the early European studies on the mathematical geography of the Islamic culture and had himself presented in several works seem not to have attracted any attention, above all, among historians of geography of the 20th century. The judgement given by Hermann Wagner towards the end of the 19th century on the historical views of his Italian colleague Gustavo Uzielli, namely that he had made a giant leap from Ptolemy to Toscanelli without considering the work of the Arab scholars of the Middle Ages, cannot be assessed highly enough. This kind of time leap from Ptolemy to the early or late Renaissance is practically typical of the geographical-historical orientation taken even as late as in the 20th century. The 800-year development phase which the discipline experienced in the Arab-Islamic civilisation is almost completely ignored. The tremendous development of mathematical geography in the Islamic world, which was already largely known in European research in the 19th century, was not considered in the debate on important questions of the history of cartography. Those who are led to consider an influence exerted by Arab-Islamic geography for toponymical reasons generally tend to seek the reasons for this in sources of a descriptive nature.

At some time or another, an approach began to pervade the work and studies of researchers into the history of cartography, which consisted of solely explaining the extremely complicated cartographic depiction of regions, such as the eastern coast of Africa, Madagascar, the Indian subcontinent or South-East Asia, as a consequence of Marco Polo’s actual or supposed contact, or that of other persons, with these areas. Such an approach is unrealistic and unsuitable. This needs to be replaced by the conception that within the Islamic culture and from the 13th century onwards, a methodologically constantly improved and spatially far-reaching mathematical geography and its related cartography led to an ever more accurate registration and depiction of the known world. In this respect, it is not possible to acknowledge Sinologist Walter Fuchs highly enough for expressing his conviction, once he had recognised the Arab depiction of the triangular shape of Southern Africa on a Chinese map dating from 1320, that the cartographic heritage of the Arabs had only been incompletely preserved for us and that their cartographers had not always reproduced the most up to date experience gained by their mariners. Essentially Fuchs is right. The field of mathematical geography, which was already actively promoted in the first third of the 9th century by Caliph al-Maᵓmūn, in the course of time and starting from a particular development stage, became very familiar to the population of the Islamic world which stretched from the western fringe of the old known world through to South-East Asia. The persons or groups of persons who took it upon themselves in this period to mathematically explore the areas closest to them were not always Muslims and were not required to be either. It was merely decisive that they carried out their work in line with the tradition which had been created in the Islamic culture. When some of the maps surviving from the 13th or 14th centuries provide us with a practically perfect depiction of the Caspian Sea and even with a rather successful depiction of Northern Asia or also of the Indian peninsula, then, in my opinion, there can be no other acceptable explanation. This view applies to the cartography of the whole Indian Ocean with its islands and the bordering coast of Africa, where an astronomical nautical science had developed over centuries with the ability to measure any transoceanic distance or degree of longitude at sea, before the Portuguese sailors appeared there and found an already matured cartographic depiction of the whole region.

There is no lack of historical evidence with which this conception can be confirmed. Historians, geographers and astronomers of the Arab-Islamic world often supply us with interesting and important clues. This is supported by the very extensive and comprehensive numerical material contained in the geographical tables. We find comparatively little support in the field of the surviving maps. Yet, even here, there are enough examples which can help our view assert itself.

In connection with the above-mentioned opinion voiced by Walter Fuchs, I would like to mention another impression that I have gained. The time span in which parts of the Earth surface were mathematically registered in the Arab-Islamic region was very large, running to around 800 years. The areas which either had been registered or were reviewed or were newly surveyed extended from the western to the eastern edge of the known world and from around 15° or 35° south of the Equator through into most northern parts of Asia. The results of the work performed in the individual geographical regions remained, sometimes for a longer period of time and sometimes even for ever, enclosed within an area, and did not reach other areas. Those who performed the work were certainly not always or not fully aware of the significance of the results they achieved. After all, even the geographers did not hear of all the results and could, over and above this, not always judge their consequences for the history of geography. There were, in all probability and as some examples have shown, intentions and attempts to record the state of knowledge that had been achieved in one comprehensive geographical work and to illustrate this knowledge in a general map. However, no attempt of this kind had even any halfway prospect of succeeding in producing a complete picture. Perhaps the world map by the Ottoman cartographer and mariner Pīrī Reᵓīs, which has been lost except for a small section, was one of these attempts as can be taken from the legend of the surviving section. And perhaps we are not wrong to see the “Javanese” map, which has been incompletely preserved in 28 sections, and which fell into the hands of the Portuguese in around 1511, as a further attempt to record and communicate the very latest results of the cartographic depiction of the Earth’s surface as comprehensively as possibly, shortly before the Portuguese made their appearance in the Indian Ocean.

The good fortune of being able to record the present status of the cartographic image of the Earth acquired by humankind in a complete depiction remained a matter for posterity, namely the 19th and 20th centuries. But even an only roughly adequate consideration and acknowledgement of the enormous share contributed by the Arab-Islamic culture to this common heritage still waits to be tackled.