![Sample our Arts journals, sign in here to start your access, latest two volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/110799/TOC-Subject-Sample-2021-Arts.jpg"})
ABSTRACT
In 1282, the Persian scholar, Quṭb al-Dīn al-Shīrāzī (d. 1311), included in a cosmographical work a description of a ‘Greek’ map of the Mediterranean. Rather than reproducing this map graphically, he placed a square grid over the image, listed the cells covered by the sea, and then itemised the toponyms according to their cell numbers. Once this pixelated map is created as he instructed, it is obvious that the ‘Greek’ map must have been an Italian portolan chart, featuring its unique elements: a realistic outline for the northern coast of Africa, especially the Gulf of Sirte, a simplistic rectangular form for the British Isles, a counterclockwise tilt, and a large size. With a toponymy that matches that on the oldest surviving portolan chart (the Carte Pisane, c.1270) and the next survivor (the Cortona chart), assigned to later that century, the ‘Greek’ map in Shīrāzī’s possession in 1282 seems to be the second oldest portolan chart of which we have specific knowledge. This article recounts how a Persian cosmographer disseminated a cartographic image based on the direct observations of Christian pelagic mariners. This unexpected material evidence confirms that knowledge of those earliest accurate sea charts was not confined to the Latin world, even if such information exchanges were probably rare.
La carte textuelle de la Méditerranée de Quṭb al-Dīn al-Shīrāzī (1282) et son évidente source dans une carte portulan
En 1282, le savant persan Quṭb al-Dīn al-Shīrāzī inclut dans un ouvrage cosmographique la description d'une carte ‘grecque' de la mer Méditerranée. Au lieu de reproduire graphiquement cette carte, il place une grille carrée sur l'image, énumère les cases couvertes par la mer, puis détaille les noms de lieux en fonction de leur numéro de case. Lorsque l’on redessine selon ses instructions cette carte composée de points, il apparaît de manière évidente que cette carte “grecque” devait être une carte portulan italienne, avec ses éléments caractéristiques : un contour réaliste pour la côte nord de l'Afrique, en particulier le golfe de Syrte, une forme rectangulaire simplifiée pour les îles britanniques, une inclinaison de son axe dans le sens inverse des aiguilles d'une montre, et un grand format. Avec une toponymie qui correspond à celle de la plus ancienne carte portulan conservée (la Carte Pisane, vers 1270) et celle de la suivante (la carte de Cortone), datant de la fin du XIIIe siècle, la carte “grecque” en possession de Shīrāzī en 1282 semble être la deuxième plus ancienne carte portulan dont nous ayons une mention explicite. Cet article explique comment un cosmographe persan a diffusé une image cartographique basée sur les observations directes de marins chrétiens. Cette source matérielle inattendue confirme que la connaissance de ces premières cartes marines précises n'était pas limitée au monde latin, même si de tels échanges d'information ont été probablement rares.
Un mapa textual del mar Mediterráneo de Qutb al-Din al-Shirazi (1282) y su fuente evidente en una carta portulana
En 1282 el erudito persa Qutb al-Din al Shirazi incluyó en un trabajo cosmográfico una descripción de un mapa ‘griego' del Mediterráneo. Más que reproducir este mapa geográficamente, colocó una cuadrícula sobre la imagen y numeró las celdas que cubrían el mar y luego detalló los topónimos según el número de su celda. Una vez que creado fue obvio que el mapa ‘griego’ debió haber sido una carta portulana italiana, presentando elementos únicos: un esquema real del contorno de la costa norte de África, especialmente el golfo de Sirte, una esquemática forma rectangular de las Islas Británicas, una inclinación contraria a las agujas del reloj y un gran tamaño. Con una toponimia que coincide con el más antiguo portulano conocido (la carta Pisana c.1270) y el siguiente que ha sobrevivido (la carta de Cortona) datada después de ese siglo, el mapa griego en posesión de Shirazi de 1282, parece ser el segundo portulano más antiguo del que tenemos un específico conocimiento. Este artículo relata cómo un cosmógrafo persa diseminó una imagen cartográfica basada en observaciones directas de marineros pelágicos cristianos. Esta evidencia material inesperada confirma que el conocimiento de estas cuidadas cartas marinas no estaba confinado a un mundo marino, incluso si estos intercambios de información probablemente fueran raros.
Quṭb-ad-Dīn aš-Šīrāzīs aus Text bestehende Karte des Mittelmeeres (1282) und ihr offensichtliches
Vorbild in einer Portolankarte 1282 fügte der persische Gelehrte Quṭb-ad-Dīn aš-Šīrāzī die Beschreibung einer „griechischen“ Karte des Mittelmeeres in ein kosmographisches Werk ein. Statt die Karte graphisch zu kopieren, legte er ein quadratisches Gitter über die Vorlage, listete die von Meer bedeckten Zellen auf und fügte eine Liste der Ortsnamen mit der Nummer ihrer jeweiligen Zelle bei. Rekonstruiert man diese „gepixelte“ Karte nach seinen Angaben, wird klar, dass die „griechische“ Karte eine italienische Portolankarte gewesen sein muss. Sie weist eindeutige Elemente auf, wie den realistischen Umriss Nordafrikas (insbesondere der Großen Syrte), die simplifizierte Rechteckform der Britischen Inseln, eine Drehung gegen den Uhrzeigersinn und ein großes Format. Da sich die Ortsnamen mit denen der ältesten erhaltenen Portolankarte, der „Carte Pisane“ (ca. 1270), und der nächsten, der „Cortona Karte“ (datiert auf das Ende des Jahrhunderts), decken, scheint die „griechische“ Karte im Besitz Šīrāzīs die zweitälteste Portolankarte zu sein, von der wir detailliertere Kenntnis haben. Der Beitrag beschäftigt sich mit der Frage, wie ein persischer Kosmograph ein Kartenbild verbreitete, das auf den Beobachtungen christlicher Seefahrer beruht. Diese überraschende Quelle belegt, dass die Kenntnis dieser ältesten genauen Seekarten nicht auf die christliche Welt beschränkt war, selbst wenn solche Informationstransfers wahrscheinlich selten stattfanden.
Acknowledgements
The authors would like to thank the anonymous referees for their constructive suggestions for improving this article. The first author would also like to thank Prof. Jamil Ragep and Sajjad Nikfahm-Khubravan for prompting the first phases of this project, Osama Eshera for generously offering invaluable feedback on various iterations of this article, and Hasan Umut for helping with acquiring manuscripts from libraries in Türkiye.
Disclosure statement
no potential conflict of interest was reported by the authors.
Notes
1 We have transliterated only the toponyms that Shīrāzī reported from the ‘Greek’ map; for all other names, including bodies of water, we have used the modern spelling/names.
2 Although mediaeval sources usually consider a parasang to be approximately three Arabic miles, the value of the mediaeval parasang or Arabian mile cannot be precisely stated in modern units.
3 The Red Sea.
4 Waystation is the distance a caravan can take in a day.
5 Strait of Gibraltar.
6 The Aegean Sea.
7 The Adriatic Sea.
8 This is probably owing to a misplacing of the diacritical points of the word Bunṭus, which is the Arabized version of Pontus, the Greek name of the Black Sea. One should note that the orthography of the words Nīṭush and Bunṭus, without the diacritical points, is the same in Persian and Arabic.
9 The Black Sea.
10 Bosporus.
11 Roughly Morocco and the western part of Algiers today.
12 Roughly Tunisia and the western part of Libya today.
13 Strait of Gibraltar.
14 Cf. Strabo’s discussion of the existence of an isthmus at the Pillars of Heracles and his rejection of Eratosthenes’ conjecture that “the breaking of the channel at the Pillars of Heracles had not yet taken place and that in consequence the Mediterranean Sea, since it was of a higher level, joined the exterior sea at the isthmus and covered it, but after the breaking of the channel took place at the Pillars, the Mediterranean Sea was lowered and thus exposed the land about Casium and Pelusium, as far as the Red Sea.” Strabo, Geography, vol. 1, ed. Jeffrey Henderson, English tr. Horace Leonard Jones (based on the unfinished version of J. R. Sitlington Sterrett), The Loeb Classical Library, 49, (Cambridge, Mass.: Harvard University Press, reprint 2005), 141.
15 By ‘this gulf,’ he means the Mediterranean Sea. Shīrāzī here refers to it as a gulf because, in mediaeval geography, it was considered to be one of the gulfs extending from the Encompassing Sea into the inhabited world.
16 In this part Shīrāzī reads the cells of the Mediterranean from Gibraltar to its utmost eastern coast in the form of vertical clusters, leaving the coverage of the Adriatic and Aegean seas for the following paragraphs.
17 Again, he means the Mediterranean Sea.
18 The Adriatic Sea.
19 These two cells had already been designated as sea cells in the previous paragraph as part of the cluster of cells from (22, 5) to (22, 16).
20 Sea of Azov.
21 The North Sea.
22 ‘(6, 13) to (6, 29)’ was dropped in all manuscripts of the Ikhtiyārāt that we have examined except one: Süleymaniye, MS Ayasofya 2574, copied two centuries after the original composition. It was later added in the margin in some early manuscripts of the Tuḥfa (Süleymaniye, MS Ayasofya 2587, Leiden, MS Or. 192 and Majlis, MS Shūrā 6130), which were copied during Shīrāzī’s lifetime, and found its way to the body of some other early ones, such as, Central Library of the University of Tehran, MS Adabiyāt 89 and Millī, MS 19040.
23 Here, again all manuscripts of the Ikhtiyārāt that we have used, except one (Süleymaniye, MS Ayasofya 2574), have the extra phrase ‘from (8, 15) to (8, 29) and from (8, 16) to (8, 29),’ which seems redundant. Two early manuscripts of the Tuḥfa (Süleymaniye, MS Ayasofya 2587 and Millī, MS 19040) do not have this seemingly incorrect phrase. A later reader of the Tuḥfa in Majlis, MS Shūrā 6130, writing in the margin of f. 107b, rectified this mistake and tells us that he made this correction “according to the grid (jadwal).” This scholium suggests that some readers of Shīrāzī not only drew the map according to his instructions but engaged his text in a critical manner.
24 Starting from this paragraph we only mention Shīrāzī’s transliteration of the toponyms. Modern and portolan equivalents of each mainland toponym are provided in . For modern equivalents of inland toponyms, names of islands, and provinces/regions respectively see , , and .
25 All along the north African coastline from here, except for the Gulf of Tana and Bernic, Shīrāzī only gave horizontal coordinates, probably because the coastal outline was already marked by the blanket sea coverage. So, the vertical coordinate can be inferred from the sea coverage.
26 Shīrāzī has ‘then Lakharaz, then Marsā’ but apparently, they should together represent Marsā al-Kharaz.
27 Shīrāzī’s use of the word ‘above,’ here and in its two next occurrences, indicates that he considered the orientation of the ‘Greek’ map to be southward. In fact, in today’s northward maps, Mahdiya is below (or south of) Tūnis.
28 Lit., gulf.
29 As explained in note 27, Shīrāzī’s use of the word ‘below,’ here and in its next occurrence, is an indication of his considering the ‘Greek’ map to be southward. In today’s northward maps, Ṭulmaythā is above (or north of) Barnīq.
30 Lit., cape.
31 The Mediterranean.
32 This is where the Battle of La Forbie took place in 1244 and was known to the Crusaders as ‘Forbie.’ See, Yāqūt al-Ḥamawī, Muʿjam al-buldān (Beirut: Dār Ṣādir, 1995–96), 4: 246, where he mentioned this name with a slightly different orthography and suggested a different spelling, Farbiyā, as a village within the administrative jurisdiction of ʿAsqalān.
33 The Mediterranean.
34 Lit., castle.
35 Ancient Mount Tmolus and modern-day Bozdağ, in Türkiye.
36 Turkmens of the frontier areas in Anatolia in the 12th–13th centuries. For more information on Udj Turkmens see Speros Vryonis, The Decline of Medieval Hellenism in Asia Minor: and the Process of Islamization from the Eleventh Through the Fifteenth Century (Berkeley: University of California Press, 1986), 133, 136, and 191–192.
37 See paragraph [9–10] above where he explains the sea coverage of cell (30, 16).
38 Inland Kastamonu.
39 One cell away from the sea.
40 One cell away from the sea.
41 He means the other side of the Kerch Strait.
42 In the Tuḥfa: Dhunawī. Idrīsī has it as Danū and Ibn Bībī as Dūnāb. For Ibn Bībī’s spelling see: Muḥammad Javād Mashkūr, Akhbār-i Salājiqa-yi Rūm, (Tehran, 1971), 298.
43 Previously defined as a water cell.
44 Previously defined as a water cell.
45 Lit., land of Gyrfalcons, which might refer to Norway. For the important role of gyrfalcons in the Middle Ages, see Thierry Buquet, “The Gyrfalcon in the Middle Ages, an Exotic Bird of Prey (Western Europe and Near East),” ed. Charles Burnett and Baudouin Van Den Abeele, Falconry in the Mediterranean Context During the Pre-Modern Era, 9, Droz, 79–98, 2021, Bibliotheca Cynegetica, 978-2-600-06236-7, accessible online: https://hal-normandie-univ.archives-ouvertes.fr/hal-02139381/document
46 King Edward I of England sent gyrfalcons as gifts to Arghūn in a diplomatic mission that left England in 1291, led by Geoffrey of Langley, see Lockhart, “The Relations,” 27. Buscarello de’ Ghizolfi (see note 14 of the main text) accompanied Geoffrey of Langley in this mission.
47 We are not sure what is being referred to by Urkīna. It sounds like Orkney, an archipelago north of mainland Scotland, comprising over seventy islands, with the largest known as ‘Mainland,’ but it is unlikely that Shīrāzī referred to them as a single island.
48 Romans here means Byzantines.
49 Since Shīrāzī considered the orientation of the ‘Greek’ map to be southward, in our northward reproduced map, Iqrītish, or the modern Crete, is indicated by colouring the upper halves of (27, 10) and (28, 10) in light grey.
50 Cephalonia.
51 All Ikhtiyārāt manuscripts supply an orthography that can be dotted to be read as Māyuṭ. This orthography is close enough to the Persian/Arabic orthography of Mālṭ for us to maintain the latter as our preferred reading and the former as an early textual corruption.
52 All Ikhtiyārāt manuscripts supply an orthography that can be read as Mardāniyā. As before, this orthography is close enough to that of Sardānyā to justify the latter as our preferred reading and to view the former as an early scribal error in the transmission history of the Ikhtiyārā.
53 Shīrāzī’s use of ‘below’ here indicates that he considered the orientation of the ‘Greek’ map to be southward. In fact, in our northward maps, Kūrsikā (Corsica) is above (or north of) Sardānyā (Sardinia).
54 All Ikhtiyārāt manuscripts supply an orthography that can be read as Qābis. This too can be justifiably read as Yābis.
55 All Ikhtiyārāt manuscripts have (36, 15). This cannot be correct because, with this coordinate, Sīwās would not be across from Dimashq, which is in (37, 9). Indeed, in all manuscripts of the Tuḥfa that we consulted, the coordinate of Sīwās has been correctly given as (37, 15).
56 Seine River.
57 Lit., city of the Pope.
1 For transliterating Arabic and Persian words, we used International Journal Of Middle East Studies (IJMES) transliteration system: cambridge.org/core/services/aop-file-manager/file/57d83390f6ea5a022234b400/TransChart.pdf. Shīrāzī was born to a family of physicians in 1239 in Shiraz (modern-day Iran). At the age of fourteen, after his father’s death, he started working as a physician and ophthalmologist at a hospital in Shiraz. After ten years, Shīrāzī abandoned that post and travelled to Khurasan in order to find a scholar with whom to study and improve his knowledge of medicine. There, he also studied philosophy and astronomy with the most renowned scholars of his time. He then went to Qazvin (also in modern-day Iran) where he studied law. Later, in Maragha (also in modern-day Iran), Shīrāzī continued studying advanced mathematical sciences and astronomy. After travels in Khurasan, central Iran, and Iraq, and further studies in grammar, philosophy, law, and other disciplines, Shīrāzī settled in Anatolia for almost twenty years. His time in Anatolia covered an even broader range of activities: from studying, teaching, and writing different works to serving as a judge and leading a diplomatic mission to the Mamluk court in Cairo. In 1290, he left Anatolia for Tabriz, where he resided until his death in 1311 and was mostly busy teaching and writing. Through his extensive travels and encounters, Shīrāzī gained access to a wide variety of sources, some of which were rare, even at his time. Shīrāzī’s works are especially significant because they preserve some works that have not otherwise reached us. See Fateme Savadi and Osama Eshera, 'Quṭb al-Dīn al-Shīrāzī's Commentary on Ibn Sīnā's Qānūn in Historical Context: Edition and Translation of the Preface,' Intellectual History of the Islamicate World (forthcoming). For a more detailed biography of Shīrāzī, see Fateme Savadi, ‘The Historical and Cosmographical Context of Hayʾat al-arḍ with a focus on Quṭb al-Dīn Shīrāzī’s Nihāyat al-idrāk’ (PhD diss., McGill University, Institute of Islamic Studies, 2018), 20–36.
2 For this eyewitness account, see Rashīd al-Dīn Faḍl Allāh al-Hamadānī, Jāmiʿ al-tawārīkh, ed. Bahman Karīmī, 2 vols. (Tehran: Shirkat-i nisbī-yi Ḥāj Muḥammad Ḥusayn Iqbāl va shurakā, 1959), 2: 822–823. The author of this chronicle, Rashīd al-Dīn Hamadānī (d. 1318), is a well-known statesman and historian of the Ilkhanid era. Please note that all translations from Persian and Arabic in this article were provided by Fateme Savadi.
3 See Savadi, ‘Historical and Cosmographical Context,’ 42–43, where she briefly explains how Shīrāzī transmitted the content of a map of the Mediterranean in Ikhtiyārāt III.1, offers a reproduced version of the map following Shīrāzī’s instructions, and points to its possible portolan provenance. An edition of the Ikhtiyārāt is under publication: Amir-Mohammad Gamini, ed., Quṭb al-Dīn Shīrāzī’s Ikhtiyārāt-i muẓaffarī (Tehran: Iranian Institute for Philosophy, forthcoming). The Tuḥfa has not been edited. For a brief survey of the geographical part of Tuḥfa III.1, see Jean-Charles Ducène, ‘Les sources de la partie géographique du Tuḥfat al-shāhiya fī l-hayʾa de Quṭb al-Dīn al-Shīrāzī,’ in eds. M. Reinkowski, M. Winet, and S. Yasargil, Arabic and Islamic Studies in Europe and Beyond, Etudes arabes et islamiques en Europe et au-delà (Leuven: Peeters, 2016), 327–344. Ducène, also, referred to Shīrāzī’s presentation of a map of the Mediterranean during his audience with the Ilkhanid ruler of Persia. It seems Ducène realised that Shīrāzī was in fact using a map, while writing the geographical part of Tuḥfa III.1, especially in his description of the Mediterranean, because he says, ‘Il situe souvent une ville par rapport à une autre en précisant que celle-ci fait face à celle-là, il précise aussi la forme de certains golfes. Ces réflexions ne sont possibles que si on a un modèle graphique en face des yeux’ (330). However, Ducène did not recognize Shīrāzī’s explicit reference to a map of the Mediterranean Sea before starting his description of this sea in the aforementioned part of the Tuḥfa, nor did he notice Shīrāzī’s transmission of this map in a textualised format later in the same part of the work. As for the composition date of the Tuḥfa, it was preserved in the authorial colophon that was transmitted in the authentic manuscripts of the work. In the case of the Ikhtiyārāt, the only known material evidence of its composition date is an Arabic note written in red at the end of one of its manuscripts by the anonymous scribe of the main text (Milli, Shīrāzī, Ikhtiyārāt, MS 31402, f. 201b; for a reproduced image of this note, see a).
4 For more information about this genre of cosmographical writing, see F. J. Ragep, Naṣīr al-Dīn al-Ṭūsī’s Memoir on Astronomy (al-Tadhkira fī ʿilm al-hayʾa) (New York: Springer-Verlag, 1993), 1:29–36; and F. J. Ragep, ‘Astronomy,’ in Encyclopaedia of Islam, THREE, ed. Kate Fleet et al. (Brill, updated to 2023), accessed 28 February 2023, <http://dx.doi.org/10.1163/1573-3912_ei3_COM_22652>.
5 For more information about the geographical content of Book III of this genre of cosmographical writing, see Savadi, ‘Historical and Cosmographical Context,’ 53–92.
6 In the Ikhtiyārāt, Shīrāzī identified his sources saying: ‘the configurations [of the seas] as visualised, formulated, and presented by others, such as the authors of the works of routes and realms (arbāb-i ṣuḥuf-i masālik va mamālik) and cosmography (aṣḥāb-i kutub-i hayʾat), among others’ (Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, fol. 138a). It is noteworthy that both the verso and recto sides of fol. 138 in this codex were transcribed by Shīrāzī himself. ‘Routes and realms’ refers to a genre of geographical writing in the premodern Islamic world. The works belonging to this genre shared a common aim and structure: geographical description of major cities, provinces, and the network of routes that connect them. The earliest extant exemplar of this genre is Kitāb al-masālik wa-l-mamālik (‘book of routes and realms’) by Ibn Khurdādhbih (d. c. 913). Later works of ‘routes and realms’ featured regional schematic maps. Two exemplars of this kind are al-Iṣṭakhrī’s (d. c. 962) al-Masālik wa-l-mamālik (‘routes and realms’) and Ibn Ḥawqal’s (d. after 978) Ṣūrat al-arḍ (‘image of the Earth’). For a discussion of the formation of the ‘routes and realms’ genre and its influence on the geographical content of the cosmographical works in the premodern Islamic world, see Savadi, ‘Historical and Cosmographical Context,’ 45–71.
7 Regarding the description of the seas, Shīrāzī was dismissive of his sources: ‘upon reflection, I found their thoughts baseless and their efforts useless, because most of the configurations [of the seas in their works] are such that their truth cannot be attested by reason or by transmitted knowledge’ (Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, fol. 138a).
8 Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, fol. 138a.
9 Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, fols 138a–138b.
10 Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, fol. 138b.
11 In the Tuḥfa, Shīrāzī states that it is difficult to reproduce the map on regular folios, due to the map’s large size. Bibliothèque nationale de France (hereafter BnF), Arabe, Shīrāzī, Tuḥfa, Sivas, 1285, MS 2516, f. 75b. In order to have a sense of the map’s size, consider the fact that each bifolio of the aforementioned BnF copy of the Tuḥfa (Arabe MS 2516), which was copied under Shīrāzī’s supervision, measures approximately 25×20 cm, as against the 70×100 cm of a typical portolan chart. In the section preceding the description of the Mediterranean Sea in the Ikhtiyārāt, Shīrāzī said that he did not reproduce the map of the Indian Ocean and its gulfs because the copyists would distort the map after copying it once or twice and mislead the naive reader. Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, f. 142b.
12 In both Persian and Arabic, Rūm ambiguously refers to Anatolia, Byzantium, and Roman lands.
13 For example, Shīrāzī cited Ptolemy’s Geographia in Ikhtiyārāt III.1 for the southern limit of the inhabited world being 16°25′ below the equator. Topkapı Saray, Ahmet III, Shīrāzī, Ikhtiyārāt, MS 3310, f. 137a.
14 For more information about de’ Ghizolfi and his diplomatic missions between the Christian and Islamic worlds, see L. Lockhart, ‘The Relations between Edward I and Edward II of England and the Mongol Īl-Khāns of Persia, ’ Iran 6 (1968): 23–31.
15 George Sarton, Introduction to the History of Science, 3 vols. (Baltimore: Published for the Carnegie Institution of Washington by the Williams & Wilkins Company, 1927), II: 775.
16 De’ Ghizolfi’s mission was to convey Arghūn’s interest in allying with the Christians against the Mamluks. Shīrāzī himself was sent by the previous Ilkhanid ruler, Aḥmad Tegüder on a diplomatic mission to the Mamluk court. De’ Ghizolfi’s mission was therefore especially sensitive, and the Ilkhanid historical sources are silent about it.
17 See respectively paragraphs [6], [7], [8], [9], [10], [11], and [13] of the Appendix.
18 Shīrāzī returned later to refine the sea coverage by specifying cells for some of the islands.
19 Youssouf Kamal, Monumenta cartographica Africae et Aegypti, IV (1936–1937), reprinted by Fuat Sezgin (Frankfurt am Main: Institut für Geschichte der Arabisch-Islamischen Wissenschaften, 1987). For his transcription and French translation of the relevant passages of the Tuḥfa and his reproduced map, Kamal used Universiteit Leiden, Shīrāzī, Tuḥfa, MS Or. 192. By contrast, our reconstruction is based on fourteen manuscripts of the Ikhtiyārāt and eleven manuscripts of the Tuḥfa. This explains the discrepancies between Kamal’s reconstruction and ours. Kamal did not add any comment about Shīrāzī’s map or its provenance, but he must have assumed that its provenance was a portolan-chart, since he included it in a volume dedicated to those.
20 J.E. Kelley, Jr., ‘Non-Mediterranean Influences that shaped the Atlantic in the early Portolan Charts,’ Imago Mundi 31, no. 1 (1979): 18–35, 21, Fig. 1. It should be noted that on page 19 and in the caption of Fig. 1 on page 21, Kelley mistakenly mentioned Shīrāzī’s Nihāyat al-idrāk as the source of the map.
21 For the Cortona chart, see Vera Armignacco, ‘Una carta nautica della Biblioteca dell'Accademia Etrusca di Cortona,’ Rivista Geografica Italiana 64 (1957): 185–223.
22 For an in-depth study of portolan chart origins see Tony Campbell, ‘Mediterranean Portolan Charts: Their Origin in the Mental Maps of Medieval Sailors, Their Function and Their Early Development,’ January 2021, https://www.maphistory.info/PortolanOriginsMENU.html.
23 See Tony Campbell, ‘Cartographic Innovations by the Early Portolan Chartmakers,’ 2016, http://www.maphistory.info/PortolanChartInnovations.html .
24 The Carte Pisane—whether from lack of space or for some other reason—omits the right third of the Black Sea, going no further east than Crimea in the north and laliminia and simiso (Samsum) along the south coast. The manner in which the toponymy is placed at the right-hand edge of the vellum shows that there had not been any trimming.
25 No doubt because of their limited relevance to Mediterranean sailors. In another major error, the mis-sizing of the Adriatic on charts continued until 1330 but the pixels are too large for that to be clear.
26 Tony Campbell and Michael Barritt, ‘The Representation of Navigational Hazards: The Development of Toponymy and Symbology on Portolan Charts from the 13th Century onwards,’ The Journal of the Hakluyt Society (December 2020), 1–36, accessible online: https://www.hakluyt.com/downloadable_files/Journal/Campbell%20Barritt.pdf.
27 The earliest surviving Greek portolan chart dates from 1459 and uses Roman lettering; the first with Greek lettering was from the sixteenth century. See George Tolias, The Greek Portolan Charts, 15th-17th Centuries: A Contribution to the Mediterranean Cartography of the Modern Period (Athens: Olkos, 1999), 179, 190.
28 Ramon Pujades, Les cartes portolanes: la representació medieval d’una mar solcada (Barcelona: Institut Cartogràfic de Catalunya; Institut d’Estudis Catalans; Institut Europeu de la Mediterrània; Lunwerg, 2007), 350–397.
29 ‘Literacy or Illiteracy?’ in Campbell, ‘Mediterranean Portolan Charts.’
30 That is not to deny that the basic portolan chart could not be augmented with other, non-marine materials, as can be seen, for example, on the Avignon and Lucca chart. See Jacques Mille, La carte d’Avignon—De la Méditerranée à la Baltique 1190–1490 (Espelette: Pyrénées-Atlantiques, 2021); Philipp Billion, ‘A newly discovered fragment from the Lucca Archives, Italy,’ Imago Mundi 63, no. 1 (2011): 1–21 and coloured plates 1–2.
31 Even though Saffi is recorded in Lo compasso de navegare (see the note immediately below), this toponym’s inclusion on the ‘Greek’ map is significant because it is highly unlikely that Shīrāzī saw that textual guide to the Mediterranean, and nowhere as far south as Saffi is recorded on any portolan chart until 1318 (i.e. Vesconte’s middle period, and more than three decades after the ‘Greek’ map).
32 On Lo compasso de navegare, see Patrick Gautier Dalché, Carte marine et portulan au XIIe siècle: le "Liber de existencia riveriarum et forma maris nostri Mediterranei" (Pise, circa 1200) (Rome: École française de Rome: distributor, Paris: Boccard, 1995); Alessandra Debanne, Lo compasso de navegare. Edizione del codice Hamilton 396 con commento linguistico e glossario (Brussels: Peter Lang for the Gruppo degli italianisti delle Università francofone del Belgio, 2011).
33 Tony Campbell, ‘Presence of Red Names on the portolan charts (up to 1400),’ 2013, http://www.maphistory.info/RedNameAnalysisTo1400.doc.
34 The breakdown of how the ninety coastal toponyms on the pixelated map are treated on the Carte Pisane is as follows: 48 in red, thirty in black, nine in areas not covered (though present on the Cortona chart) and just three missing.
35 BnF, Arabe, Shīrāzī, Tuḥfa, Sivas, 1285, MS 2516, f. 76a.
36 Here, to have a sense of the number of names missing between Tripoli and Alexandria, it would be helpful to mention their corresponding Campbell toponymical listing numbers, respectively: 1651 and 1556. For convenience, in Tables 1, 2, and 3, the toponyms are provided with their numbers in a comprehensive Excel listing (Campbell numbers), arranged in a clockwise geographical sequence: Tony Campbell, ‘Listing and analysis of portolan chart toponyms along the continuous coastline from Dunkirk to Mogador (early 14th to late 17th century) including the transcribed names from the Liber de existencia riveriarum and Lo compasso de navegare as well as the Carte Pisane and Cortona chart,’ 2015, http://www.maphistory.info/PortolanChartToponymyFullTableREVISED.xls.
37 The Campbell toponymical listing number for sarabiom is 1614. Before portolan-chart toponymy began to be standardised by Pietro Vesconte after 1313, each of the charts drawn prior to that exhibited a number of idiosyncrasies, including several unique names. See Table E in Tony Campbell, ‘Comparison of the “Precursor” and “Antecedent” Names on Two 13th-century portolani and Four Supposedly Very Early Anonymous Charts,’ 2015, http://www.maphistory.info/PisaneRed&BlackNamesTABLES.doc. See also Tony Campbell, ‘Listing of the 35 names found, in red, only on one or more of the four supposedly “early” anonymous charts,’ 2013, http://www.maphistory.info/RedNamesEarliestChartsRareUnique.doc.
38 See note 32.
39 The Carte Pisane is illegible at the relevant point.
40 See in Tony Campbell, ‘Carte Pisane Hydrography Tables,’ 2015, http://www.maphistory.info/CartePisaneHydrographyTables.docx.
41 The occasional suggestion that portolan charts were an Islamic invention has been generally dismissed. See Pujades, Les cartes portolanes, 508–509. The earliest surviving Arabic examples date from no earlier than the first half of the fourteenth century, and then clearly as copies of Christian charts. Indeed, the fact that Shīrāzī went to considerable trouble to reproduce the example he acquired demonstrates that Islamic examples of such documents could not have been available at that time. The earliest reference to a Muslim sailor using a portolan chart appears in a Mamluk encyclopaedia entitled Masālik al-abṣār fī mamālik al-amṣār (‘routes of insight into the realms of cities’) by Ibn Faḍl Allāh al-ʿUmarī (d. 1349). In his description of the Mediterranean Sea, ʿUmarī used a portolan chart that he acquired from a Muslim sailor, called Abū Muḥammad ʿAbd Allāh ibn Abī Nuʿaym al-Anṣārī al-Qurṭubī. According to ʿUmarī, Qurṭubī (lit. ‘from Cordoba’) dedicated his youth to sailing in the Mediterranean Sea. Qurṭubī taught ʿUmarī about the portolan chart and its rhumb lines. ʿUmarī then described and reproduced the rhumb lines in a section on winds in his encyclopaedia. It should be noted that ʿUmarī was active approximately twenty years after Shīrāzī’s death, meaning about fifty years after Shīrāzī’s composition of the Ikhtiyārāt. ʿUmarī’s citations of a portolan chart are discussed and translated to French in: Jean-Charles Ducène, ‘Le portulan arabe décrit par al-ʿUmarī,’ Cartes et géomatique 216 (2013): 81–90.
42 See note 7.
43 The earliest reference appears in Guillaume de Nangis’s account of an incident on Louis IX’s voyage of 1270. For the 1279 Egidio Colonna reference, see Emmanuelle Vagnon, ‘Cartographie et représentations occidentales de l’Orient méditerranéen, du milieu du XIIIe à la fin du XVe siècle,’ Terrarum Orbis 11 (2013): 140–142. That the ‘Liber de existencia riveriarum’ (probably before 1210) took its distance and direction statements from a marine chart, makes it an indirect reference.
44 Jacques Mille, ‘The Avignon Chart (c.1300–c.1310): An Early Attempt to Represent the Northern Regions of Europe on a Nautical Chart,’ Imago Mundi 75, no. 2 (2023). See also, Mille, La carte d’Avignon.