Islamic Civilization

Cartography and Earth Representation

The development of Muslim cartography highlights and confirms all previous observations made, that the development of Muslim sciences was in response to practical necessities; that Muslims corrected their predecessors through repeated experiment, calculation, and the use of instruments, and in the end, Islamic contributions were the foundations of our modern discoveries, including in this case maritime and other discoveries.

The need for maps was imposed on Muslims by conditions pertaining to the historical situation, which followed the rise and early advance of Islam in the late 7^th-early 8^th century. Tibbetts notes how, prior to the establishment of Islam in the 7^th century, there is no evidence that the early Arabs accepted the idea of representing landscape in a systematic way.[1] Toward the end of the first century of Islam, we do find a few literary references to military maps. A map was prepared ca 83 H/702 of the country of Daylam, south of the Caspian Sea, for al-Hajaj Ibn Yusuf (d. 714), the governor of the Eastern Province so that the military situation could be understood from his capital in Iraq .[2] Similarly he ordered a plan for Bukhara  so that he could acquaint himself with the layout of the city  while preparing to besiege it in 707.[3] A map of the swamps of al-Batiba near Basra  was also said to be available in the time of the Abbasid Caliph, al-Mansur, this time because of a dispute over fresh water.[4] A plan was also reputedly drawn in 758 for the round city of Baghdad  planned for Caliph al-Mansur. Since this was a strongly walled fortress in which only the privileged lived, the drawing was less a city plan than an architect’s site plan,[5] but since the city had a diameter of two kilometres or more, the plan may have been a substantial cartographic attempt.[6] 

Another instance of the ruler involving himself in the scientific drive, for whatever practical purpose,  was clear from the Surat al-Ard (Picture of the Earth) map, prepared under the Caliph al-Mamun. Al-Mamun’s map is, indeed, likely to have been prompted as much by a political motive as by a purely scholarly one.[7] The map is not extant, but according to al-Masu’di who had seen it, it was in various colours, and depicted the universe with the spheres, the stars, land and the sea, inhabited and barren regions of the earth, settlements of people, cities, etc.[8] Al-Masu’di also insists that this map was better than anything that preceded it.[9] Al-Zuhry (d. 1154? 1161?) claims that his own Jughrafya was based on a copy of al-Fazari’s  Jughrafya, which itself was based on the Jughrafya of Caliph al-Mamun, in whose elaboration more than seventy scholars had participated.[10]       

Muslim map makers initially relied on their Greek predecessor, Ptolemy, but, soon, contrary to what is generally asserted, the Muslims deviated considerably  from Greek geography, which became in their hands completely  redundant. As Bagrow observes, whilst Muslims held his work, the Almagest in high esteem, they soon forgot his Geographia.[11] And the view that Muslim cartography ‘turned back the clock when it broke away from Greek traditions represented by Ptolemy, is unfounded,’ Bagrow insists.[12] The Muslims had early felt the need for accurate determination of position, if for no other reason than that the correct orientation of their mosques depended upon such knowledge.[13]  As Koyre points out, Muslims took on the role as the precursors of the art of precision and accuracy.[14] Every Muslim scientist gave himself to the improvement of methods of finding these geographical coordinates, and with commendable results: their figures were a considerable improvement on those given by Ptolemy.[15] Tangier as measured from Alexandria, for instance, was not 53⁰30’ but 35⁰41’; Ptolemy’s measurement of the Mediterranean  was 19 degrees in excess.[16] Moreover, as noted by al-Biruni, places east were found in the west, and vice versa.[17]

Amongst the earliest to make amendments to Greek geography, and contribute a more accurate knowledge and mapping of the earth was Al-Khwarizmi  (d. 863). His Surat al-Ard[18] (Image of the Earth) is a rich treatise on Mathematical geography.[19] It corrects and fills gaps left by Ptolemy.[20] Al Khwarizmi’s  tables cut down Ptolemy’s estimate of the length of the Mediterranean  Sea from 62 degrees to about 52;[21] a length which was further cut by Al-Zarqali of Spain.[22] Al-Khwarizmi’s Surat al-Ard, more importantly, signals a new form of geography, where the world is in the form of a sphere.[23] It is divided in its northern half into seven longitudinal bands from the equator: the iqlim, reviewed and corrected according to new data and knowledge, with latitudes and longitudes locating the principal mountains, rivers, seas and towns on the picture.[24] It gives latitudes and longitudes for 2402 localities as a basis for a world map, and also includes latitudes and longitudes for cities, islands, and geographical regions.[25]

Al-Zarqali gives the length of the Mediterranean  Sea in his commentary on the Toledo  tables as 42 degrees, which is almost exact (Ptolemy gives 62 degrees), whilst al-Marrakushi gives a list of 135 coordinates which, for accuracy, deserve to be regarded as one of the most important contributions of Muslims to mathematical geography.[26] These and other Muslim contributions led to more precise location of places, old and new, and contradict sharply widely held opinions by modern historians, such as Parry, who asserts that: ‘The scholars of early fifteenth century Europe had no reliable criteria for criticising Ptolemy.’[27]

In order to calculate the terrestrial latitude and longitude of places, the Muslims developed a number of techniques and methods. The method of finding latitudes is here stated by Al-Battani :

‘If you want to know the latitude of any given town, then, first of all, measure the altitude of the sun at noon (i.e when it crosses the meridian) in any direction. When you know the altitude then, find out the declination of the Sun at that time, if it is north then deduct it from the altitude, and if it is south, add it to the altitude. This gives the altitude of Aries or Libra at that town. Subtract this from 90^0,  the remainder will be the latitude of the town.’[28]

As regards longitude, which is the angular distance east or west from the standard meridian to that of any place, all they had to do was to find out at any given instant, the time and the standard meridian (Muslim geographers generally adopted the Fortunes islands as the standard meridian; others adopted the meridian of Arin (a place in India )), as well as the local time at the station concerned, and then, by converting the difference of the time into degrees they obtained the longitude of the place.[29] They faced, of course, the problem of finding the correct local time, which they eventually solved by means of the eclipse of the moon because it supplied the requisite information, i.e the time at any given instant at the prime meridian as well as the local time at the station concerned.[30] So Muslim geographers collected from various quarters the local times at which the eclipse occurred, and from that data, they easily calculated and verified the longitudes of the places of observation.[31]

Al-Biruni  made further contributions towards more accurate measurements in setting the foundations for modern mathematical geography.[32] He carried geodetic measurements and determined with remarkable precision the coordinates of a number of places, besides introducing a simplified method of stereographic projection.[33] In his works he gives the theory behind two different projections, one of which would be known today as an azimuth equidistant projection and the other as a globular projection.[34] Al-Biruni also criticises the projections of Ptolemy and Marinus.[35] Al-Biruni mentions in some detail the measurements of the degree of latitude.[36] He carried this out in Khwarizm and in Ghazna, and he produced a new method of measurement by using a convenient mountain from which the horizon could be observed.[37] Al-Biruni also sought to measure the difference in longitude between two places using the distance between them in miles.[38] This was difficult since the direct distances between places could not be worked out with accuracy; however, he produced a result for the longitude of Ghazna east of Baghdad , setting out the theory behind this operation so that it was there for any later scholar to improve.[39] Al-Biruni also took it for granted that the earth is round, noting "the attraction of all things towards the centre of the earth," and remarked that astronomic data can be explained as well by supposing that the earth turns daily on its axis and annually around the sun, as by the reverse hypothesis.[40]

A further impetus to the development of cartography took place in the 10^th century,  when  a school of geography rose and exerted a decisive impact on the subject.[41] This school’s distinctive feature was that texts and maps went together, each text accompanied by a set of twenty-one maps: one for each of the twenty climes of the Islamic lands, and one world map.[42] Members of this school were al-Balkhi and his followers: al-Istakhri, Ibn Hawqal, and al-Muqaddasi.[43] The inspiration of this school was al-Balkhi (born Ahmed B. Sahl, about 849-50) near Balkh from which he took his name al-Balkhi.[44] The son of a schoolmaster, he travelled as a young man to Iraq , met the celebrated al-Kindi, and under the latter's influence became enthusiastic about the exact sciences.[45] In  920 or a little later, he wrote Kitab al-Ashkal or Suwar al-Akalim (The Book of the Forms and the Images of the Continents) a text accompanying maps of the inhabited world, which was unusual then.[46] The first map is a map of the world, circular in form, and Makkah  its centre; the world is surrounded by the ocean, and two gulfs enter the continent. These gulfs are the Mediterranean  and the Indian Ocean ; Arabia  is seen as the centre of the world, and next North Africa , Muslim Spain, Egypt , and Syria .[47] The second part of the geographical description deals with eastern Islam.  Although lost, this work impacted on those, which followed.[48] Amongst these is Al-Istakhri’s (fl. C. 950) Kitab al-Masalik wal Mamalik (Book of Journeys and Kingdoms) which describes lands he visited, each with a chapter and a map in colour.[49] Coast-lines and rivers are represented under conventional forms, many of al-Istakhri’s maps showing the Mediterranean in a circular or elliptical form.[50] Ibn Hawqal (fl 977) drew 22 provincial maps of the Islamic realm, amongst these maps is one of the Nile, reproduced by Kramers in his new edition of Ibn Hawqal.[51] From a look at the maps, it would seem quite obvious that the best representation of the Mediterranean, its shores, islands, and location of cities, amongst the early geographers is that of Ibn Hawqal.[52] Al-Muqaddasi was the last representative of this school, creating, according to Tolmacheva, the systematic foundation of Muslim geography by discussing its uses and scopes, the geographic terminology, the various methods of division of the earth, and the value of empirical observation[53] He was in the company of the senior sea captains, the mathematicians, the middlemen and the merchants, who had the clearest perception about the Red Sea, its ports, winds, islands, and he enquired from them about its nature and limits, their records he studied carefully, making suitable selections from them, comparing them to maps he had seen earlier.[54] Al-Muqaddasi was the first to produce maps in natural colours. He says:

‘In the maps we have coloured the familiar routes red, the golden sands yellow, the salt seas green, the well known rivers blue, and the principal mountains dull brown.’[55]

Al-Muqaddasi also seems to indicate the relative importance of the towns by the differing sizes of circles, something he is very keen on in his text.[56] In all three respects, introducing coloured maps, different colours for rivers, routes, etc, and differing sizes for urban areas, Al-Muqaddasi clearly innovated, and set up a methodology future geography was to adopt.

The medieval culmination of geography was probably reached with al-Idrisi (b.1101). He spent most of his working life at the court of the Normans in Sicily . The Sicilian ruler, Roger II was his patron, and thanks to his position, Al-Idrisi was able to acquire information about France, Italy, Germany and parts hitherto missing in previous Muslim works.[57] A typical instance of his knowledge of current events is supplied by the account of the Lisbon ‘wanderers’,[58] three men who went on an exploratory mission in the Atlantic Ocean, an early (9^th century) adventure, some of whose details have reached us.[59] This, in addition to its intrinsic value as being the earliest recorded voyage after the rise of Islam, deliberately undertaken for the purpose of exploring the Western ocean, provides an interesting spotlight upon the breadth of Al-Idrisi’s outlook.[60] Al-Idrisi’s final output was what is commonly known as al-Kitab al-Rujari (The Book of Roger), which includes descriptions of the world as known then; accompanied by a spherical world map, and a larger rectangular map. The spherical map was made of a silver ball of about 400 kgs, on which were drawn the seven continents, lakes and rivers; mountains and plains, cities and routes (and distances, heights, or lengths for all),[61] with great accuracy.[62] The silver planisphere is lost (probably melted in the course of the centuries). Bagrow, closer to the historical truth, says that the ‘silver tablet,’ fell in the hands of the mob and was smashed to pieces in the year 1160;[63] the year 1160 being when widespread massacres of Muslims, including of courtiers, took place.[64] The large rectangular map, however, features in many works. It consists of seventy regular sheets assembled in seven rows of ten sheets each, and a small circular world map. Against today’s conventions, north was at the bottom, and South on top. On this map are located hundreds of cities and detailed information of all sorts.[65] By this map recent studies of particular regions were inspired.[66] Al-Idrisi’s map corrected many once held medieval assumptions including the perception in Europe that the world, apart from Europe, belonged to the Muslims.[67]

Western historians claim that both Islamic geography and cartography are a reproduction of Greek maps, and worse, that knowledge of the world and the seas by the time of the Renaissance had not evolved since Ptolemy, as here held by one leading Western historian, Parry,  who claims:

‘In the process of Reconnaissance, explorers by sea, pushing rashly out in the world of the unknown, but for Ptolemy.’[68]

This is false. In an admirably conceived appendix by Maqbul, we come across a good comparison between Ptolemaic maps and al-Idrisi’s, which show how the latter made a through revision of nearly everything that Ptolemy did.[69] This is succinctly outlined here.

* Ptolemy placed the northern limits of the known world at 63⁰ N. Lat. And the southern limits at 16⁰ 25’’ S. Lat., the total north-south extent of the known world being 79⁰ 25’.[70] Al-Idrisi on the other hand, extended the northern limits to 64⁰ N. Lat. without specifying the southern limits.[71]

* Al-Idrisi divided his world map between the equator 0⁰ and 64⁰ N. Lat. Into 42 parallels. But the distance between one parallel and the other varies, and even within the seven climes, drawn as straight lines, the distance is not equal. Assuming that the parallels are equidistant from each other, the distance between each parallel would be approximately 1⁰ 35’ 25’’. According to Ptolemy who drew 21 parallels on his map, the distance between each parallel was 3⁰. Hence the ratio as given by Ptolemy and al-Idrisi between each parallel would be 1:2. South of the Equator, al-Idrisi added space on his map equal to about the distance between two parallels. So, the total north-south distance of al-Idrisi’s map (which is 79.3 cm) would be 69⁰ 58’ 20’’ as compared to Ptolemy’s 79⁰25’. Thus, al-Idrisi reduced the size of his map by 9⁰ 26’ 40’’. This resulted in shifting the equator by 3⁰ 37’ north. Since al-Idrisi included on his map all the regions shown by Ptolemy as well as some additional countries of the Baltic Sea, the geographical positions of places were changed by him specially in the northern and the southern regions.[72]

* In his extension of the eastern coast of Africa, al-Idrisi also differed from Ptolemy. Since al-Idrisi reduced the scale of his map, north south, the eastern coast of Africa on his map is shown turning east at approximately 4⁰ Lat. N. where he places the equator, whereas according to Ptolemy, it turned eastwards at 15⁰ Lat S; 80⁰ Long. or at the point where he placed the Prasum Promonotry. This resulted in the reduction of the size of the Indian ocean on al-Idrisi’s map.[73]

* With regard to the Seas, Al-Idrisi counted seven seas, six of which are connected with each other by water whilst another, the Caspian, was landlocked. These seas were the Indian Ocean , the Mediterranean , the Gulf, the Red Sea, the Black Sea, the Adriatic, and the Caspian.[74] Ptolemy counted three such seas: the Mediterranean, the Caspian, and the India  Sea.[75]

* On the Indian Ocean , al-Idrisi’s map reduces the size of this ocean in comparison to the size given by Ptolemy, who considered it the largest of all seas.[76] According to Ptolemy, it was surrounded by land on all sides, but on al-Idrisi’s map, its eastern stretches are shown connected with the Pacific between 1⁰ S. Lat. and 4⁰ N. lat., approximately, but in the text he describes its conjunction with the Pacific at 13⁰ Lat.S.[77]

* The shape of the Caspian Sea is changed by al-Idrisi in relation to that drawn by Ptolemy; in Ptolemy’s it is shown longer west to east than north to south, whereas al-Idrisi shows it longer north to south than east to west.[78]

* Al-Idrisi introduced considerable modifications in Ptolemy’s map of Europe, the shapes of some of the countries and their coastlines were changed. In many cases, the names, provincial and regional divisions were replaced by new ones. Idrisi also extended the northern limits of the inhabited world from Ptolemy’s 63⁰ N. lat to 64⁰ N. lat. Al-Idrisi adds countries such as Poland, Norway, Sweden, etc, which are all absent in Ptolemy.[79]

* In drawing the map of Asia, Al-Idrisi changed almost the whole map of Asia, both in respect of the Ptolemaic political boundaries as well as its physical features.

* Al-Idrisi also adds to the knowledge of places, rivers, mountains, etc, of Asia, Africa, and Europe, and also political divisions such as in North Africa . Ptolemy with regard to the latter speaks of Mauritania, Tringitania, etc, Al-Idrisi replaces them with contemporary names and divisions, and also introduces the names of the settlements of the African tribes.[80]

To assume, as the majority, if not all Western historians, that al-Idrisi’s maps did not impact on the Christian West is not just fallacious, but is also proof of poor scholarship. For, how could a Muslim geographer (Al-Idrisi) be brought to a Christian court, the leading sea-going power of the time, and create for this court first class mapping, just for his work then to be discarded. This makes no sense. It is not because modern historians have, for whatever reason, failed to show how the Christian West adopted Idrisi’s knowledge that such adoption did not take place. It is impossible for such knowledge not to be adopted, for al-Idrisi’s knowledge remained valued and valid for centuries. Le Bon, for instance, points out that Al-Idrisi’s reference to the sources of the Nile was only matched by European discoveries in the late 19^th century.[81] More important is a detail,  may be short, but highly crucial, stated by Bagrow in passing:

‘In 1154, a few weeks before Roger’s death, manuscripts of the book (Al-Idrisi’s work) in Latin  and Arabic were completed, together with the map, seventy sheets…’[82]

The focus here is on the word Latin , which escaped so many, for, of course, the work by al-Idrisi did not need any translation, for it was in Latin from the outset (he was working for a Christian ruler, after all), and that’s why it was easy to conceal its impact, for no medieval Western scholar needed to translate it. So, rather than being overlooked, as we read in modern Western historians’ writing, this work was simply available, ready for use from the source. 

A further major Islamic contribution to mapping was the greater use of instruments in search of ever more accurate calculations and locations. Amongst the instruments used are the astrolabe, the ruba (quadrant), the gnomon, the celestial sphere, the sundial, and the compass.[83] Al-Khujandi (fl second half of 10^th century), for instance, built a universal instrument al-Ala al-Shamila (the comprehensive instrument) which was used instead of the astrolabe or the quadrant, but it could be used for one latitude only. Al-Badi al-Ustrulabi al-Baghdadi (d. 1140) constructed an astrolabe which could be used for all the latitudes.[84] Finding the latitude of any place by the use of this astrolabe can be outlined here:

‘Take the reading in the meridional line as shown in almuqantara (the circle of latitude). This space equals the height of the Arctic pole from the horizon of that place. If you desire to know the latitude of the region, take the altitude of the sun in the middle of the day when the sun is in Aries or Libra. Subtract the degrees of the sun’s altitude from 90, the remainder will be the latitude of the place. Supposing that the sun on that day at noon was 38⁰ 10 in elevation, the latitude will be 51⁰ 50’.’[85]

The impact of such instruments on better mapping was obvious. Muslim maps, Bagrow remarks, show ‘artistic and ingenious schematic drawings’; compasses, ruler and set square producing the necessary geometric figures.[86]

Considering the 12^th century translations from Arabic into Latin  and other languages,[87] contemporary Latin works derived from the Islamic originals, such as John of Holywood’s Tractatus de Sphaera, or Sphaera Mundi, completed in 1233,[88] and other forms of Islamic influence in the use of instruments,[89] prompts the conclusion that there was a much better and more accurate knowledge of the earth at the dawn of the age of the great discoveries. As Kramers points out, in Islamic maps there are no pictures of men or animals (which is a reflection of the Islamic ban on the representation of either), which make Christian maps appear more fantastic,[90] but far more unrealistic.

The pioneering Islamic role in map making has only been recently recognised in one of the best works on the subject: Harley and Woodward’s History of Cartography, a significant endeavour; of which volume two, book one is devoted to Islamic cartography.[91] The richness of information is such that to even seek to outline the points of interest raised by the various contributors is unrealistic. Subjects vary from celestial mapping, to the history of Islamic cartography, geodesy, pre-modern Ottoman mapping, marine charting, and so on and so forth. And from the first page of the preface of this volume, one gleans a valuable comment:

‘Even in narrative discussions of the history of cartography, such as those of Lloyd Brown and Gerald Crone, and nothwistanding the universal ring of their titles, the non European mapping traditions were largely ignored.[92] Such an approach in the standard texts taught several generations of students that the history of cartography was largely a Western achievement and part of the history of European science. Quoting an Islamic historian of science, it was as if the descent of maps had passed ‘directly from the Greco-Roman period to the European Renaissance as if nothing took place in the history of science and technology from the fall of Rome in the late 5^th century to the fall of Constantinople in the fifteenth.’[93]

Harley and Woodward have thus put their finger on the problem that this work has battled with from the first page, that is the disregard for the Islamic contribution to science and civilisation that has plagued mainstream history of science.[94]

One key matter that has generally received no interest in the study of geography, despite its fundamental importance, is the Muslim representation of the spherical shape of the earth. Both Dreyer and Kramers have raised this issue.[95] Kramers, for instance, states, that the Muslims had no objection to the spherical form of the earth, then denied by Christian theologians.[96] Thus, an early Muslim geographer, Ibn Rusta (fl. 903), says:

‘The stars which are familiar in the northern skies gradually sink lower in the heavens. The Constellation of the ‘Great Bear,’ which in our skies never sets, does set when sufficient southern latitude is attained, on the other hand constellations new to the inhabitants of the northern climates are seen to rise above the southern horizon. Another proof of sphericity of the earth is furnished by the fact that the moon and the stars do not rise at one and the same time in all parts of the world, as we see them rise in the east before their rise in the west, similarly their disappearance takes place earlier in the east than in the west.’[97]

More importantly, in the world of Islam, Kramers holds, there is no record of any Muslim being persecuted for stating that the earth was a sphere, and that it was also very small compared to the size of the universe.[98] In the Christian West, before the year 1000, Rybka points out, intellectual standards were extremely poor even amongst the supposedly well read, and this hardly encouraged the penetration of the advanced sciences of the Islamic world.[99] At that time, in the Christian West, there prevailed the naïve idea derived from a simplistic interpretation of the Bible, that the earth could not be a sphere.[100] This was not the case in the Muslim world, where this spherical shape was obvious in the maps the Muslims drew, and also in the globes and spheres they built. Earth is often represented in its spherical shape, most of its surface made of water, the inhabited parts constituting only a fraction of the whole surface, and deserts occupying a good place.[101] This Islamic breakthrough, neglected today, did, however, shatter one of the most established views of the time that the earth was flat. This breakthrough opened the way in Western Christendom to the challenge of established norms, a crucial development in intellectual history. The spherical shape of the earth also raised the possibility that one end of the world could be reached by sailing in an opposite direction. The impact of this is well known. Without it, as Kramers recognises, the discovery of America would have been impossible.[102] And this leads onto the next matter: nautical science.