The Nautical Atlases of cAlī al-Sharafī
Mónica Herrera-Casais
Key-words: cAlī al-Sharafī, Sharafī family, Maghrebi and Andalusian
chartmakers (A¬mad al-Æanjī, Ibrāhīm al-Mursī and A¬mad al-Andalusī),
Majorcan sources, nautical atlas, Ðabla (tabula), sea chart of the world,
jughrāfiyā, al-Idrīsī’s Geography, coastal place names, wind network, dry
point, pomegranate, moriscos.
Abstract
The two extant atlases of the 16th-century Tunisian chartmaker and scholar
c
Alī al-Sharafī are indispensable documents for understanding the history
and development of Maghrebi nautical cartography. They provide
precious information on al-Sharafī’s chartmaking techniques, production
and sources, which include the output of his own family and the atlas of a
certain A¬mad al-Andalusī who had settled in Istanbul. The making,
contents and characteristics of the atlases are examined in connection with
the Maghrebi and Majorcan traditions of nautical cartography which
positively influenced al-Sharafī’s compilation of coastal place names.
This is explained by the circulation of such cartography across the
Mediterranean of his time. Further questions are also raised on the
working conditions in which Maghrebi and Andalusian chartmakers
flourished and the impact of their legacy.
Contents
I. INTRODUCTION: 1: The chartmaker al-Sharafī. 2: The Sharafī family
of chartmakers. 3: Al-Sharafī’s sea charts of the world. 4: Other Maghrebi
Suhayl 8 (2008) pp. 223-263
224
M. Herrera-Casais
and Andalusian chartmakers and their output. 5: The chartmaker A¬mad
al-Andalusī.
II. THE ATLASES: 6: Presentation. 7: Codicological description. 8: Terminology: al-Ðabla. 9: Contents and composition.
III. THE SECTIONAL CHARTS: 10. Geographical distribution and
coastline layout. 11. Coastal place names. 11.1: In the Islamic
Mediterranean. 11.2: In Mediterranean Europe and the Black Sea. 12: The
wind network. 13: The pomegranate. 14: Conclusion. 15: Acknowledgements. 16: Bibliography.
I. INTRODUCTION
1. The chartmaker al-Sharafī
c
Alī al-Sharafī (fl. 1551-79) is the best known and the most popular of all
Maghrebi chartmakers, with some thirty years of activity. 1 He belongs to
a larger family of chartmakers from Sfax, an Islamic corsair port in
Tunisia, from where he eventually moved inland to Kairouan. 2
Apparently, it was here that he finished one of his nautical atlases in 1571
and one of his sea charts of the world in 1579 which he signed as
...al-¼afāqusī mansha’ an wa-mawlidan al-Qayrawānī qarāran wa-maskanan
al-Mālikī madhhaban (from Sfax by origin and birth, from Kairouan by
settling and residence, al-Mālikī by legal doctrine). Al-Sharafī’s two
nautical atlases from 1551 and 1571, 3 and a sea chart of the world from
1579 are his only extant works. Three other comparable charts can be
safely attributed to him, according to a passage in his second atlas. Still,
1
His cartography has aroused great interest among modern scholars, such as Nallino 1944
and Soucek 1992, pp. 284-87, 289, figs. 14.20-25 (b/w). For general references, see GAL,
Suppl. 2, p. 710; Ma¬fūÞ 1982-86, vol. 3, no. 281; MGC, vol. 12, pl. 7c-d (b/w) and vol.
13, pp. 430-31.
2
At the beginning of the 16th century, the main industry in Kairouan was based on leather
crafts for trade with Numidia: see Leon l’Africain: Description de l’Afrique. Translation
by A. Epaulard, Paris, 1956, vol. 2, p. 398; quoted by M. Talbi: “al-°ayrawān,” EI2 4
(1978). The parchment obtained from animal skin was fundamental to Mediterranean
chartmakers, as it was the material on which most sea charts and many atlases were made:
see Astengo 2007, p. 182ff.; Campbell 1987, p. 376.
3
Henceforth, for the sake of convenience, these will often be designated as the first (1551)
and second (1571) atlases.
The nautical atlases of cAlī al-Sharafī
225
repetitive features of his works suggest a regular production of such
atlases and charts, which might have allowed him to earn a living as a
professional chartmaker. The contents of the atlases attest to al-Sharafī’s
mapmaking skills, but also to his notions on medieval geography and folk
astronomy, a knowledge which is not documented for every
Mediterranean chartmaker.
2. The Sharafī family of chartmakers
The Sharafī family, originally from Sfax, ran the only recorded Maghrebi
workshop in which chartmaking skills were probably taught by father to
son, imitating the same kind of cartography for more than fifty years. The
Sharafīs are widely known for their characteristic sea charts of the world
which combine Mediterranean nautical cartography with al-Idrīsī’s maps.
Their extant output, though incomplete, represents the greater part of the
surviving tradition of Maghrebi chartmaking. This includes another sea
chart of the world by al-Sharafī’s son (Mu¬ammad), dated 1600-01. All
the family charts, including al-Sharafī’s lost exemplars, were apparently
derived from a prototype designed by his grandfather.
In the Mediterranean, family workshops were not restricted to the
Sharafīs, and had already flourished in 14th-century Majorca with the
Jewish chartmakers Abraham Cresques (fl. ca. 1375) and his son Jafudā
(fl. ca. 1400), already professionally active. 4 Between the 16th and 17th
centuries, we find the prolific families of Prunes/Pruners, established in
Majorca, and Oliva/Olives, of Majorcan origin but spread mainly in
Marseille, Naples and Messina. 5 The cartographic models of Majorcan
4
The former is believed to have been the author of the Catalan Atlas: see the latest edition
as L’Atles Català (2005); and the extensive description of it by Sáenz 2007. Distinctive
features of this atlas appear in other anonymous sea charts which have been therefore
attributed to the Cresques’ workshop. These are compiled by Pujades 2007, C15-16,
18-20, 22 (with a commentary on p. 256ff.); see also Rosselló 2000. On Jafudà’s activity
in particular: R. Skelton: “A Contract for World Maps at Barcelona, 1399-1400,” Imago
Mundi 22 (1968), pp. 107-13. Moreover, Campbell 1987, p. 428ff. discusses the working
techniques of 14th-15th-century Majorcan and Venetian workshops. Astengo 2007, pp. 18991 deals with the same topic for the later period up to the 17th century.
5
Their production is listed by Rey & García 1960, pp. 45-46, 95-100, 119-63; and recently
by Astengo 2007, appendix 7.1 (commentary on p. 207ff.). See the specific studies by G.
de Reparaz: “Els Prunes, cartògrafs catalans dels segles XVI i XVII. A propòsit d’uns
mapes inèdits d’aquests cartògrafs de Mallorca,” Estudis Universitaris Catalans 13 (1928),
pp. 324-402; and Rosselló 1995, pp. 39-41 (“Les cartes i els atles dels Olives”). See also
226
M. Herrera-Casais
workshops would eventually acquire financial value. This factor is
documented by an affidavit from 1625, according to which Pere Joan
Pruners (fl. 1625-51) had received in inheritance half of the models
(patrons de fer cartes de navegar) belonging to his father, Vicenç Pruners
(fl. 1597-1609). 6 This poses the question whether the models in
possession of the Sharafīs had ever gained financial worth.
The Sharafī family produced at least four consecutive generations of
chartmakers:
1. The grandfather Mu¬ammad al-Sharafī al-¼afāqusī (fl. beginning of the
16th century), perhaps the earliest chartmaker in the family, is known to
have made a sea chart of the world, apparently lost. His grandson (cAlī)
and great-grandson’s (Mu¬ammad) exemplars are based on this model.
2. The father is A¬mad b. Mu¬ammad al-Sharafī al-¼afāqusī (fl. first half
of the 16th century). His activity (shughl) as a chartmaker is suggested
by his son cAlī al-Sharafī in his 1571 atlas (see the Arabic passage in
section 5), who also says to have learned from him about the
determination of prayer times by means of shadow lengths.
3. cAlī b. A¬mad b. Mu¬ammad al-Sharafī al-¼afāqusī al-Qayrawānī
al-Mālikī made at least two nautical atlases in 1551 and 1571, and four
sea charts of the world: three possibly earlier than 1571, which are lost,
and one from 1579.
4. The son Mu¬ammad b. cAlī al-Sharafī al-¼afāqusī is the author of
another copy of the family sea chart of the world which he signed in
1600-01. This is extant in Paris, Bibliothèque nationale de France (Rés.
Ge. C. 5089). 7
Comes 2004, pp. 555-57; Rosselló 2000, pp. 38-41, 72-74. In addition, Cartografía
Mallorquina, Barcelona, 2000 shows examples of their cartography which are kept at the
Naval Museum (Museu Marítim) in Barcelona.
6
Cf. Pujades 2007, p. 213. A Latin transcription of Vicenç Prunes’ 1609 testament is given
by G. Llompart: “Registro de los cartógrafos medievales activos en el puerto de Mallorca,”
Anuario de Estudios Medievales 27:2 (1997), (pp. 1117-48) 1127, 1146-47.
7
See Cartes nautiques sur vélin, p. 98 (no. 60); and some remarks in Nallino 1944, pp. 54041; Soucek 1992, p. 287, figs. 14.24-25 (b/w). Available reproductions: A) E.-F. Jomard:
Les monuments de la Géographie ou recueil d’anciennes cartes européennes et
orientales… publiés en fac-similé de la grandeur des originaux, Paris, 1842-62, pl. XII.
This is a sketch facsimile of the chart in which the original Arabic place names have been
The nautical atlases of cAlī al-Sharafī
227
The scholarly activity of later members of the Sharafī family is recorded
by the 18th-century chronicler Ma¬mūd Maqdīsh of Sfax (Nuzhat al-anÞār
fī cajā’ib al-tawārīkh wa-l-akhbār). 8 Among them, the most important is
the muwaqqit A¬mad b. cAbd al-Azīz al-Sharafī al-Azharī, who wrote a
treatise on the astrolabic almuqantar quadrant (Al-Durar al-fākhirāt fī
l-camal bi-rubc al-muqanÐarāt fī jamīc al-aqÐār wa-l-jihāt, comp. ca.
1735). 9
3. Al-Sharafī’s sea charts of the world
The extant 1579 sea chart of the world by cAlī al-Sharafī is kept at the
Istituto Italiano per l’Africa e l’Oriente in Rome (size 59 × 135 cm). 10
This is a copy of the prototype made earlier by al-Sharafī’s grandfather,
and consists of two sheets of parchment pasted together. The eastern sheet
covering Asia reproduces the regional maps of al-Idrīsī’s Geography
(Nuzhat al-mushtāq fī-ikhtirāq al-āfāq, comp. Palermo ca. 1154) for that
part of the world. 11 The western sheet, covering the Mediterranean, the
corrupted. It also omits the added Latin inscription on the neck of the parchment. B)
Mappae arabicae, vol. 5 offers a large b/w sketch, including a transliteration of the main
inscriptions. This is reprinted in smaller size in Pinna 1996, vol. 1, pp. 186-87. C) The
latter provides a large colour plate which is the best and more reliable reproduction of the
chart. D) Other small colour pictures appear in Itinéraire du savoir en Tunisie, p. 84; and
MGC, vol. 12, pl. 8a.
8
See Maqdīsh, Nuzha, vol. 2, p. 390ff. This source is described by C. A. Nallino: “Venezia
e Sfax nel secolo XVIII secondo il cronista arabo Maqdīsh,” Raccolta..., Roma, vol. 5
(1944), pp. 345-402.
9
See Maqdīsh, Nuzha, vol. 2, pp. 394-97, 400. He is also recorded in Cairo Survey, pp. 1056 (D48); GAL, vol. 2, p. 612 & Suppl. 2, pp. 486, 694; Ma¬fūÞ 1982-86, pp. 166 (no. 278),
170 (no. 280); Nallino 1944, pp. 541, 542; and Soucek 1992, p. 287.
10
This was first described by Nallino 1944, with an Italian translation and commentary of the
main inscriptions; see the chartmakers’ signature on p. 536. Its size is given by Soucek
1992, p. 289. A small colour picture of the chart is in The History of Cartography, vol. 2:1
(1992), pl. 24; and Z. Chelli: La Tunisie au rythme des cartes géographiques, Tunis, 1996,
pl. 22. A sketchy drawing of it appears in Atlas of Islam (1981 & 2001), pl. 1b; next to a
reconstruction of al-Idrīsī’s image of the world (pl. 1a).
11
On al-Idrīsī’s cartography, see Maqbul 1992, pp. 156-74; Mappae arabicae (K. Miller),
vols. 1, 6; idem: Weltkarte des Arabers Idrisi vom Jahre 1154, Stuttgart, 1928; and
228
M. Herrera-Casais
Black Sea and a section of the Atlantic (including the British Iles and
south Scandinavia), is based on a Majorcan sea chart initially consulted by
the grandfather. The use of this Majorcan source is stated in one of
al-Sharafī’s inscriptions which says: Io ho copiato questo mappamondo
(jughrāfiyā) da un altro disegnato da mio nonno Mu¬ammad –che Dio ne
abbia misericordia– il quale aveva copiato le coste del mare Siro ed i suoi
porti da una carta nautica (qunbā½) fatta dalla gente di Maiorca –che
Iddio la stermini–. 12 Moreover, a passage in the 1571 atlas informs us that
by that time al-Sharafī had finished three other sea charts of the world,
equally based on al-Idrīsī’s Geography. These charts, otherwise unknown,
are described as follows:
: أ ﻮا وأ ﺎ
ﺎل آﺎ اﻷ ﺮف و ﺎ هﺬ ا ﺔ و ّ اﷲ ﺎ ﻰ وأ
ﺰهﺔ
ﺎ
ﺎ ذآﺮ
آ ﺎر ﻰ13 و ﺪ ﺮز ّ ﻼث ﺮا ﺎت
ا ﺮاق ا ﺎق ﻬﺎ ﺔ اﻷرض وذآﺮ اﻷ ﺮ ا ﺬآﻮرة آّﻬﺎ و ﺎ
ا ﺎق
و ﻬﺎت آ ّ ﻬﺎ ﻰ.ﺎل و ﻮن وأود ﺔ و ﺪن ﻬﻮرة
اﻷرض
ا ﻬﺎت ا ﺮو ﺔ ﺪ ا ﺮؤ ﺎء ّﻰ أ ّﻬﺎ ﺎ ﺮ ﻬﺎ ﺮ ﺎ و ﺮ ﺎ و ﻮ ﺎ
ي
ّ أ
أو ﻬﺮ أو ﺮآﺔ أو
ُ ﺮف ﻬﺎ آ ّ ﺪ ﺔ أو
و ﺔ
أراد ذ ﻚ
.ﺪ وآﺬ ﻚ ا ﺮا ا ﻮ ﺔ ﺎﻹر ﺎء ﻬﺎ
ا ﻬﺎت و أ
اﻷ ﺪار ﻮت أو ﺎة أو ﺮ
اﻷ ﺪي ﻰ
ﻬﺎ ﻷّﻬﺎ
ّذ ﻚ اﻷ ﻮر ا
.ﻬﺎ اﻷ ﻼك
The author of this text and manufacturer of this atlas (al-Ðabla) may God
Almighty grant him success and improve his conditions and deeds–says:
I have already produced three large world maps (jughrāfiyāt kibār) based
on the description of the author of Nuzhat al-mushtāq fī-ikhtirāq al-āfāq.
They contain a depiction of the world (½ifat al-arÅ) with the names
(dhikr) of all the known seas, mountains, springs, rivers and famous
cities. All wind directions (jihāt) are drawn in accordance with those
familiar to sea captains, to enable them to sail (li-yusāfira) east, west,
north and south, and to learn from where and in which direction they
should head for any city, mountain, river, lake or spring, as well as for
especially La Géographie d’Idrīsī: Un atlas du monde au XIIe siècle. Digital facsimile of
MS Paris, Bibliothèque nationale de France: arabe 2221. Coord by A. Vernay-Nouri, Paris,
2000.
12
Cf. Nallino 1944, p. 538.
13
MS: ﺮا ﺎت
The nautical atlases of cAlī al-Sharafī
229
renowned harbours and anchorages. Whoever wants them should look
for their whereabouts for, like all worldly belongings, they change hands,
due to death or other vicissitudes of life.
[1571 atlas, fol. 3r]
Al-Sharafī calls these three works large world maps (jughrāfiyāt kibār), 14
and hints at their utility in (Mediterranean) navigation because of their
wind directions. He also denotes this kind of chart as jughrāfiyā (lit.
geography) in the above mentioned inscription of his 1579 exemplar. 15
Curiously enough, an equivalent expression appears in a note on the back
of Gabriel de Vallseca’s 1439 chart, which describes it as ampia pele di
geografia due to its large size (75 × 122 cm). 16 In the 14th-century
encyclopaedia of Ibn FaÅlallāh al-cUmarī (Masālik al-ab½ār fī mamālik
al-am½ār), the term jihāt (pl.) is already applied to the nautical wind
directions, and their graphical representation (as rhumb lines) on
Mediterranean charts. 17 These are actually drawn in the extant sea charts
of the world by al-Sharafī and his son (1600-01), which show a network
of wind directions on each sheet of parchment. This network has a
decorative, rather than practical, utility on the sheet covering Asia. The
14
The term jughrāfiyā (sing.) is a transliteration of the Greek geography ( εω ρ φί ),
though in Arabic it was also used for a world map, see S. Maqbul: “Djughrāfiyā,” EI2 2
(1965); idem: “KharīÐa,” EI2 4 (1978), p. 1077. Andalusian and Maghrebi sources, among
them al-Sharafī (in both his 1571 and 1579 works), render the term as jacrāfiyā (changing
ghayn for cayn muhmala) instead of the standard jughrāfiyā. This is attested in the title of
al-Zuhrī’s Geography (Granada, 12th century), which is edited as Kitāb al-Jacrāfīya by M.
Hajj-Sadok, in Bulletin d’Etudes Orientales 21 (1968), see pp. 22-23 (translated by D.
Bramon: El mundo en el siglo XII, Barcelona, [1991], p. xv). This particular spelling is
recorded by R. Dozy (Supplément, vol. 1, p. 198) who quotes from Ibn Khaldūn (d. 1406)
and al-Maqqarī (d. 1632).
15
Cf. Nallino 1944, p. 538. See footnote 43.
16
This note was written on the occasion of selling the chart to Vespucci around 1480, and
reads: Questa ampia pele di geografia fu pagata da Amerigo Vespucci CXXX ducati di oro
di marco. Cf. O. López: “Col·lecció del Museu Marítim de Barcelona,” in Cartografia
mallorquina, Barcelona, 2000, (pp. 89-111), 109 (and fig. on p. 108); Rosselló 1995, (pp.
53-56, pl. 3) p. 53. On Vallseca and his works, see Pujades 2007, C40-43, 74 and p. 260.
The visual features of Vallseca’s cartography are discussed in comparison with those of
the Catalan Atlas and other 14th-15th-century Majorcan charts by Sáenz 2007.
17
See Ibn FaÅlallāh: Masālik... Facsimile reconstruction by F. Sezgin, Frankfurt a. M., 1988,
vol. 2, pp. 165-75; and a commentary in MGC, vol. 11, pp. 49-58, vol. 12, pl. 34a-d.
230
M. Herrera-Casais
close similarities between the extant charts by the Sharafīs suggest that the
lost exemplars might have shared with them comparable features in terms
of cartography, wind system, format and decoration.
The idea of producing an extended sea chart that would depict the
whole world known to the medieval scholars was not unique to the Sharafī
family. In Majorcan nautical cartography this is attested on a different
format in the Catalan Atlas and the Catalan World Map of Biblioteca
Estense (henceforth, Estense World Map) attributed to the Jewish
chartmakers Abraham Cresques (fl. ca. 1375) and Pere Rossell (fl.
1447-69) respectively. 18 In addition, it has been proposed that two charts
of Angelino Dulcert (1339) and Jehudā ben Zara (1497) were originally
larger than the existing one piece parchment covering the
Mediterranean. 19 If that be true, these charts might have had a surface
shape and extension more or less akin to the Sharafīs’ works.
4. Other Maghrebi and Andalusian chartmakers and their output
The few Arabic sea charts of the Mediterranean that are known bear
witness to the blossoming of a class of highly skilled Maghrebi and
Andalusian chartmakers who had settled in Tunisia (Tunis, Sfax and
Kairouan), Tripoli and Istanbul. The timeframe of their cartographic
output spans from the first half of the 14th century to 1600, and includes
the anonymous Maghreb Chart (ca. 1325-50). 20
18
See the modern edition of the Estense World Map (MCE 1995 in the final bibliography),
besides the studies by Pujades 2007, esp. pp. 260-61; and mainly Sáenz 2007, p. 320ff.
Other charts by Rossell are described by H. Winter: “Petrus Roselli,” Imago Mundi 9
(1952), pp. 1-11; while Rosselló 2000 discusses the chartmaker. The Catalan Atlas is
mentioned in footnote 4.
19
This hypothesis is presented respectively by R. Pujades (see his contribution to L’Atles
Català, vol. 1, p. 24ff.) and R. Almagiá (Monumenta cartographica vaticana, vol. 1:
Planisferi, carte nautiche e affini dal secolo XIV al XVII. State of the Vatican City:
Biblioteca Apostolica Vaticana, 1944, pp. 47-48, pl. XIX). It should be noted that Ben
Zara’s chart is signed in Alexandria, but belongs to the Majorcan tradition. On the
chartmaker Angelino Dalorto/Dulcert, see also Pujades 2007, C7-9, pp. 254-56; and Sáenz
2007, p. 299ff.
20
This is kept at Milan, Biblioteca Ambrosiana: SP II 259. On the Maghreb Chart, see
mainly Vernet 1962; besides Herrera 2008, pp. 287-88; MGC, vol. 11, pp. 27-31, 57, vol.
12, pl. 35 and vol. 13, p. 11ff.; Pinna 1996, vol. 2, pp. 112-23 (with a superb colour plate);
The nautical atlases of cAlī al-Sharafī
231
A¬mad al-Æanjī is the only recorded chartmaker to have lived in Tunis
before the Sharafī family. His extant sea chart of 1413-14 is a highly
sophisticated example of Islamic cartography, in terms of accuracy and
illumination. 21 Its comprehensive coverage of Arabic coastal place names
for the Islamic Mediterranean hints at the existence of other
14th/15th-century Maghrebi, and to some extent Andalusian, sea charts that
al-Æanjī could have used. These earlier sources should have laid the
groundwork of Arabic nautical toponymy and hydrographic details which
characterize the later Maghrebi chartmaking.
Ibrāhīm al-Mursī was active in Tripoli, maybe Tripoli in Syria, where a
community of moriscos had settled after the fall of Islamic Murcia in
1243. His extant sea chart is a luxurious piece dated 1461. 22 Al-Mursī is
not the only Andalusian chartmaker in the Islamic Mediterranean, for a
certain A¬mad al-Andalusī is supposed to have flourished in Istanbul (see
below). Their background raises many questions about the nature of the
Andalusian contribution to the Maghrebi (and perhaps Ottoman)
chartmaking tradition, and about their degree of exposure to the language
in which the Majorcan cartography was written.
These Maghrebi and Andalusian chartmakers were not fully
independent from the influence of Majorcan and Venetian nautical
cartography from which they borrowed visual features and place names;
the latter mainly for Mediterranean Europe and the Black Sea. Al-Æanjī
brings together elements from both Majorcan and Venetian sources, 23
Soucek 1992, pp. 263-64, fig. 14.1. Additional discussion on the dating of the chart is
offered by Campbell 1987, pp. 418, 423, 445.
21
This is kept at Istanbul, Topkapi Library: Hazine 1823, and has been described by Herrera
2008. See also MGC, vol. 11, pp. 31-32, 39, 40, 136, vol. 12, pl. 36 and vol. 13, p. 13ff.;
Soucek 1992, p. 264, fig. 14.2; and D. Uçar: “Über eine Portolankarte im TopkapiMuseum zu Istanbul,” Kartographische Nachrichten 37 (1987), pp. 222-28.
22
This is kept at Istanbul, Naval Museum (Deniz Müzesi): no. 882. See MGC, vol. 11, pp.
32-33, vol. 12, pl. 37; Soucek 1992, pp. 264-65, fig. 14.3; and M. Comes: “Ibrāhīm
al-Æabīb al-Mursī,” in Las artes y las ciencias en el Occidente musulmán, Murcia, 2007,
pp. 74-81; idem: “Barcelona a la cartografia náutica àrab” (forthcoming); and Herrera
2009.
23
According to Herrera 2008, al-Æanjī’s work shows similarities with the chart of Albertin
de Virga (1409) and the anonymous kept at Barcelona (Archive of the Crown of Aragon:
no. MP-1). These are in turn strongly influenced by the style of the Venetian chartmaker
Pietro Vesconte (fl. Venice, 1311-27). Particular elements in al-Æanjī’s chart, such as his
miniatures of Scandinavian fauna, might be inspired in Majorcan sources.
232
M. Herrera-Casais
whereas al-Mursī and the Sharafīs seem to rely more on the Majorcans. In
fact, al-Mursī’s work displays clear signs of contact with the 1456 chart
by Jaume Bertrán and Berenguer Ripoll made in Barcelona. This shows
that he probably used a contemporary model from the workshop of one
these chartmakers. 24 However, some of the Majorcan distinctive features
are conceptually derived and stylistically updated from al-Idrīsī’s
Geography and other Arabic sources. This is best illustrated by their
depiction of the Atlas Mountains in North Africa, 25 which is adopted in
al-Mursī’s chart, and later in the richly ornamented Ottoman Turkish one
of ©ājj Abū l-©asan (16th century). 26
The circulation of sea charts between the two shores of the
Mediterranean favoured the mutual influence of Muslim and Christian
chartmakers who, far from working in isolation, seem to have been aware
of the trends and improvements of each cartographic tradition. Evidence is
scarce on the quantity and characteristics of Majorcan charts that reached
the Maghrebi and Levantine ports. At any rate, their distribution was
favoured by the Arago-Catalan maritime expansion and trading network
across the Mediterranean. 27 To what extent the emigrant Andalusians ever
carried Majorcan charts with them to exile is open to speculation.
Nonetheless, we know about Jahudā ben Zara, a Sephardic Jew of perhaps
24
His source is identified by Herrera 2009. On Bertrán & Ripoll’s chart, see Pujades 2007,
C58, p. 261.
25
Cf. Comes 2004, p. 543; and Y. Fall: L’Afrique a la naissance de la cartographie
moderne: Les cartes majorquines (XIVe - XVe siècles), Paris, 1982, p. 212ff. The Majorcan
version of the Atlas Mountains is described by Sáenz 2007, pp. 394-400. In addition,
Herrera 2009 discusses al-Idrīsī’s influence on the Majorcan depiction of the rivers
Guadalquivir and Segura. See also S. Brentjes: “Revisiting Catalan Portolan Charts: Do
They Contain Elements of Asian Provenance?” In Ph. Forêt & A. Kaplony (eds): The
Journey of Maps and Images on the Silk Road, Leiden, 2008 (forthcoming).
26
This later chart shows many characteristic Majorcan features, and is preserved in Istanbul,
Topkapi Library: Hazine 1822. See MGC, vol. 11, p. 33, vol. 12, fig. 38; Soucek 1992, p.
265, fig. 14.4.
27
Already in 1354, King Pere the Ceremonious ordered all galleys under control of the
Arago-Catalan Crown to have sea charts: cf. Comes 2004, p. 525. In addition, the
merchant Domènec Pujol from Barcelona is known to have exported charts to Alexandria
by the end of the 14th century: cf. Campbell 1987, p. 437. See also Pujades 2007 on the
early production and trading of Majorcan charts.
The nautical atlases of cAlī al-Sharafī
233
Majorcan origin, who signed his three extant charts of Majorcan style in
Alexandria and Safed between 1497 and 1505. 28
Given the fact that 15th-century Tunis had been the main commercial
capital of central Maghreb, it is likely that sea charts and sailing
instructions from around the Mediterranean were accessible to al-Æanjī. In
the 16th century, the advent of the Ottoman rule in Tunisia (with the short
interval of Spanish domination over Tunis and Goletta in 1573-74) should
have created a positive atmosphere for the flow of cartography to and
from Istanbul. In fact, al-Sharafī eventually copies from the atlas of
al-Andalusī who lived in the Ottoman capital. It is also interesting to note
that his 1551 atlas was acquired in Istanbul shortly before it was sent to
the French royal library around 1739. Al-Æanjī and al-Mursī’s charts are
currently kept there.
5. The chartmaker A¬mad al-Andalusī
In his 1571 atlas, al-Sharafī states to have consulted the sea charts of an
atlas by an otherwise unknown Abū l-cAbbās A¬mad al-Andalusī, a
resident of Istanbul. Whether this source was already used for his 1551
atlas is hard to determine. Neither can we say whether it contained
additional material, such as calendars, a wind rose or a world map,
comparable to al-Sharafī’s second atlas, where we read:
ا ﺮ واﻷوراق ا ﻜ ﻮ ]ﺔ[ ﻬﺎ ا ّﺮ اّﺬي ﺪؤ
ا ﻼد وا ﺮا
وأ ّﺎ
ﺔ رأ ﻬﺎ ّ ﺾ
[]ﻬﺎ
ّ ز ﺎق ﺔ إ ﻰ ا ﺎم و ّﺮ ﺮ ا ﻜ ﺎ ﺈ
ﻬﺎ وه
ﻮل وهﻮ ا ﻜ أ ﻮ ا ّﺎس أ ﺪ اﻷ ﺪ ّ ا ﺎ
ﺈ
ا ﺎآ
ا ﻮا ﺪ وا ّﺪ ر ﻬ ﺎ اﷲ ﺎ ﻰ وذ ﻚ ﻷ ّﻬﺎ
[ﺮ ا ﺔ اّ أ ﺮ ]ﻬﺎ
. ﻬﺬ
ا ﻮ ﻮدة ﺪي
As for coastal cities and harbours, they are presented on the charts
(al-awrāq, lit. pages) that show the lands extending from the Straits of
Ceuta to Syria (Mediterranean Sea) and the lands of the Sea of al-Kafā
(Black Sea). I have copied them from an atlas (Ðabla) made by an
Istanbul resident, the learned Abū l-cAbbās A¬mad al-Andalusī, who had
settled there. This atlas is different from the work (shughl) of my father
28
See Campbell 1987, p. 436, fig. 19.18; C. Roth: “Judah Abenzara’s Map of the
Mediterranean World, 1500,” Studies in Bibliography and Booklore 9 (1970), pp. 116-20;
and M. Comes & M. Herrera, “Convivencia and chartmakers across the Mediterranean,”
2009 (forthcoming). See also footnote 19.
234
M. Herrera-Casais
and grandfather –may God have mercy upon them– but I used it as it was
the only one I had with me while making (my atlas).
[1571 atlas, fol. 13r]
The name of A¬mad al-Andalusī suggests that he might have been a
morisco, or a descendant of a morisco family, who had settled in Istanbul,
probably after the Ottoman conquest of the city in 1453. In fact, migration
waves of moriscos towards the Islamic Mediterranean intensified after the
fall of Islamic Granada in 1492, due to religious persecution in the Iberian
Peninsula. Two 16th-century note-books provide detailed itineraries for the
clandestine exile of moriscos on the way to Istanbul, with intermediate
stops at Venice and Salonica. 29 Other moriscos reached Istanbul by
crossing the Maghreb. By the early 16th century, a community of moriscos
from Aragon, Valencia and Seville had established in Galata, the former
Genoese colony of Pera across the Golden Horn. 30 They coexisted with a
large Frankish population of diplomats, merchants and mariners whose
presence certainly favoured intercultural exchange with the local ruling
Ottomans. 31 Whether al-Andalusī had truly settled in Galata or this was
the place where other Andalusian chartmakers flourished remains in the
dark.
In any event, Galata would have been a favourable enclave for the
emergence of chartmaking, profiting from the maritime activities of the
imperial Ottoman shipyard and arsenal. 32 This can be inferred from Evliyā
29
Cf. L. López-Baralt & A. Irizarry: “Dos itinerarios secretos de los moriscos del siglo
XVI,” in Homenaje a Alvaro Galmés de Fuentes, Madrid, 1985, vol. 2, (pp. 547-82), 556,
558-59.
30
Cf. Epalza 1992, pp. 283-84. The anonymous literary account Viaje de Turquía (mid-16th
century) mentions them, but focuses on the Christian captives in Galata: see the edition by
M.-S. Ortolá, Madrid, 2000, p. 765ff. (“Cómo cada día se huyen d’España moriscos y
judíos”). According to Seyā¬atnāme (vol. 1:2, p. 51), they used to gather at the so-called
mosque of the Arabs, located at the former church of Saint Paul.
31
See Seyā¬atnāme, vol. 1:2, p. 49ff.; and C. J. Heywood: “GhalaÐa,” EI2 Suppl. 5-6 (1982);
L. Mitler: “The Genoese in Galata: 1453-1682,” International Journal of Middle East
Studies 10:1 (1979), pp. 71-91; E. Dursteler: “Neighbors: Venetians and Ottomans in Early
Modern Galata.” In J. Helfers (ed.): Multicultural Europe and Cultural Exchange,
Turnhout, 2005, pp. 33-47.
32
On both places, see Seyā¬atnāme, vol. 1:2, p. 43ff.; besides the monograph by I. Bostan:
Osmanlı bahriye teşkilātı: XVII. Yüzyılda tersāne-i āmire [Ottoman Naval Organization:
The Main Maritime Arsenal in the Seventeenth Century], Ankara, 1992.
The nautical atlases of cAlī al-Sharafī
235
Čelebi’s (Seyā¬atnāme) account of the Istanbul guild parade of 1638,
which has sections on sailors of the Black Sea and the Mediterranean, as
well as all kind of workers from the Galata shipyard. Among the latter, a
group of chartmakers (e½nāf-i kharīÐaciyān) is described as parading next
to the compass (e½nāf-i puslaciyān) and hour glass makers (e½nāf-i qum
sācatçiyān). 33 They are said to have had eight map shops, with some
fifteen active chartmakers, against the twenty-five shops of compass and
hour glass makers. The latter two nautical devices apparently enjoyed a
flourishing trade also in Galata. 34
II. THE ATLASES
6. Presentation
Al-Sharafī’s oldest extant atlas is dated 1st Ramadan 958 H. (31st August /
1st September 1551 A.D.). 35 Its highly artistic execution as a work of art
indicates that by that time al-Sharafī had already been a qualified
chartmaker (figs. 1, 3). His other atlas is an extended functional version,
with a larger set of additional material and textual instructions, but less
33
Here, the Turkish term e½nāf should be interpreted as profession, rather than an organized
guild: cf. E. Yi: Guild Dynamics in Seventeenth- Century Istanbul: Fluidity and Leverage,
Leiden, 2004, p. 46. The relevant passage is in Seyā¬atnāme, vol. 1:2, pp. 131-32 (also
translated by Soucek 1992, p. 284; idem 1996, pp. 93-94). Unfortunately, chartmakers are
not mentioned in the account of Sūrnāme-i Hümāyūn (edited by G. Procházka, Istanbul,
1995) of the 1582 imperial circumcision festival in which seamanship professions already
paraded.
34
Cf. Seyā¬atnāme, vol. 1:2, pp. 53, 131-32.
35
Colour pictures of the whole atlas are to be found in Itinéraire du savoir en Tunisie
(commentary by M. Chapoutot: “Les Charfi et la cartographie,” pp. 84-95 with some
inaccuracies). Pages of this atlas often appear in publications of the Bibliothèque nationale
de France, such as L'Art du livre arabe: Du manuscrit au livre d'artiste (Paris, 2001), p.
127 (scheme of sacred geography) and La mer: Terreur et fascination (Paris, 2004), p. 23
(world map). See also E. Blochet: “Contribution a l’étude de la cartographie chez les
musulmans,” Bulletin de l’Académie d’Hippone 29 (1898), (pp. 1-27) 15-18; and Mappae
arabicae, vol. 5, pp. 150, 154-55 (with sketches of the world map and the scheme of
sacred geography) and vol. 6 (1927), pl. 78-80.
236
M. Herrera-Casais
decoration (figs. 2, 4). 36 This second work is dated at the end of jumādā
al-ukhrā 979 H. (late November 1571), some weeks after the Ottoman
defeat at the naval battle of Lepanto (October 7th). In fact, it omits
Lepanto from the sequence of coastal place names, in contrast to the first
atlas in which it appears with the Ottoman Turkish designation
Aynabakhtī*, apparently written in al-Sharafī’s hand. 37 He either
overlooked it or omitted it on purpose in the second atlas, in which case
this would acquire a political significance.
The atlases show distinctive ornamental features based on the colour
red, and frames and medallions of intertwined knots. These are also
peculiar to the sea charts of al-Sharafī and his son, as well as the earlier
ones of al-Æanjī and al-Mursī, which are all framed by a pattern of such
knots. 38 This type of design is widely documented in Islamic art and
manuscript illumination, especially in Qurans from al-Andalus and the
Maghreb. Both al-Sharafī’s atlases, mainly the first one, are often
displayed as visual icons of Islamic cartography, similarly to other
emblematic works within the context of Majorcan, Venetian and Ottoman
chartmaking.
36
Two rather problematic translations of this atlas are available by W. Brice (Manchester, ca.
1981, unpublished) and M. Mansouri (“Une famille de cartographes tunisiens: Les Sharfi,”
in H. Akkari (ed.): La Méditerranée médiévale: Perceptions et représentations,
Paris/Tunis, 2002, pp. 263-77). The latter translation omits the last two pages of the text,
and all the tables and cartography of the atlas. Some of Brice’s sketches of the cartography
(world map, scheme of sacred geography and chart of the Iberian Peninsula) and the table
of lunar mansions are reproduced in the guard leaves of the Atlas of Islam (1981 & 2001).
In addition, the Arabic incipit and colophons are edited by A. al-Ghunaym: Arabic
Geographic Manuscripts at the Bodleian Library, Oxford, 2006, pp. 298-300. Pinna 1996,
vol. 2, pp. 140-61 (with superb colour pictures) deals with the chart of the west central
Mediterranean in both atlases.
37
The asterisk (*) denotes the toponyms written in red ink. See the description of Lepanto in
Ba¬rīye, vol. 2, fols. 157r–158v, and fig. in p. 681; F. Babinger: “Aynabakhti,” EI2 1
(1960). It is registered with the name Lepanto in Majorcan and Venetian sea charts: cf.
Kretschmer 1909, p. 634.
38
On the knotted decoration of Maghrebi and Ottoman sea charts, see Herrera 2008, p. 286.
D. King: “An Astrolabe from 14th-Century Christian Spain with Inscriptions in Latin,
Hebrew and Arabic: A Unique Testimonial to an Intercultural Encounter,” Suhayl 3 (200203), pp. 9-156 draws attention to the knots engraved on Islamic astronomical instruments;
he mentions al-Sharafī’s atlases on pp. 98-101.
The nautical atlases of cAlī al-Sharafī
237
Who were the atlases made for?
Mediterranean nautical cartography is expected to have had a practical
utility, that is, to have been used as a navigational guide for sea captains
on board ship. However, the stylish decoration and artistry of most
surviving Maghrebi, Majorcan and Venetian examples, sometimes with
elements in golden ink (for instance, al-Æanjī and al-Mursī’s works)
suggest that such pieces had been more likely intended for private
collectors and armchair scholars with a general knowledge on geography
and navigation. Al-Sharafī’s atlases, in particular the visually striking
1551 version, fit into this category. This kind of public also explains the
nature of the Sharafī’s sea charts of the world, partly based on al-Idrīsī’s
Geography, still at the turn of the 17th century.
In any event, both al-Sharafī’s atlases show traces of having been used
by Ottoman readers. The 1551 atlas has added place names as well as the
names of the eight main winds, all written in Ottoman Turkish by a
different hand than al-Sharafī’s, on the charts of the Aegean and the Black
Sea. In the 1571 version, the island of Sardinia is likewise labelled
(Sardāniya) by a Turkish Ottoman hand on the chart of the west central
Mediterranean.
7. Codicological description
The 1551 atlas is preserved in the Bibliothèque nationale de France, Paris
(MS arabe 2278, 8 fols.). The catalogue of the old royal library, states that
it was acquired in Istanbul, perhaps by a French cultural expedition,
shortly before 1739. 39 Its excellent condition reinforces the idea that it
was a luxury book. This atlas is made entirely of paper paste folios
(24.5 × 20 cm), with the inside of the binding covered in Turkish marbled
paper.
The 1571 atlas entered the Bodleian Library, Oxford (MS Marsh 294,
originally 14 fols.) around 1697, as part of the manuscript collection of the
39
Cf. A. Melot: Catalogus codicum manuscriptorum Bibliothecae Regiae, pars prima, Paris,
1739, p. 189 (no. 847). See also W. de Slane: Catalogue des manuscrits arabes de la
Bibliothèque Nationale, Paris, 1883 (1st fasc.), p. 399. Several unidentified geographies are
recorded among the manuscripts acquired by the 17th-century French expeditions to the
Levant. See the inventories collected by H. Omont: Missions archéologiques françaises en
Orient aux XVIIe et XVIIIe siècles, Paris, 1902, 2 vols.
238
M. Herrera-Casais
Dublin archbishop Narcissus Marsh. 40 The atlas is in a fragile condition
indicating that it was heavily used, though not necessarily on a boat. One
folio that contained a scheme of moon phases and a calendrical table is
missing. The sectional sea charts are drawn on paper paste folios while the
rest is written on European paper with a watermark (size 27 × 21 cm).
The French orientalist Jean Gagnier (ca. 1670-1740), 41 working at the
Bodleian, translated al-Sharafī’s authorship inscription into Latin on the
front page of the atlas (fig. 2). He also added the Holy Land place names
Ascalon, Jafa, Acco, Tyrus and Said on the chart of the Levant, as well as
scattered notations on other folios. His interest in al-Sharafī’s atlas
encouraged him to make his own copies of the scheme of sacred
geography, the world map and the wind rose. These are collected in a
manuscript miscellanea of Gagnier’s writings on geography (Oxford,
Bodleian Library: MS Or. 304, fols. 1-3). 42
40
Cf. E. Bernard: Catalogi librorum manuscriptorum angliae et hiberniae in unum collecti,
Oxford, 1697, vol. 2, part 2, p. 54 (no. 1169). See also Nicoll 1821, p. 602; J. Uri:
Bibliothecae Bodleianae codicum manuscriptorum orientalium… Pars prima. Oxford,
1787, pp. 202-03 (no. 935).
41
See M. J. Franklin: “Gagnier, John,” in L. Goldman (ed.): Oxford Dictionary of National
Biography, Oxford, 2008 (electronic version at <http://www.oxforddnb.com/>).
42
See Nicoll 1821, p. 233 (MS Arab. Moh. 266).
The nautical atlases of cAlī al-Sharafī
Fig. 1: AL-SHARAFĪ’S 1551 ATLAS
Front page with authorship and dating inscription (fol. 1v).
Courtesy of the Bibliothèque nationale de France, Paris (MS arabe 2278).
239
240
M. Herrera-Casais
Fig. 2: AL-SHARAFĪ’S 1571 ATLAS
Front page with authorship and dating inscription (fol. 1r).
Courtesy of the Bodleian Library, Oxford (MS Marsh 294).
The nautical atlases of cAlī al-Sharafī
241
8. Terminology: al-Ðabla
Al-Sharafī designates his own atlases, and that of al-Andalusī, with the
term al-Ðabla. An inscription on his 1579 sea chart of the world also calls
it al-Ðabla, though al-Sharafī generally denotes this kind of map as
jughrāfiyā. 43 The word as such exists in classical Arabic, perhaps derived
from the Greek τά λ (writing support; drum), reaching the Arabs
through Middle Persian. 44 Its use by al-Sharafī attests to its value as a
technical term within the context of Maghrebi chartmaking. Here Ðabla is
contaminated by the Medieval Latin tabula, which designates both an
atlas and a map, including a sea chart. The Latin word was equally applied
to the Tabula Amalfitana, one of the oldest compendia of Mediterranean
maritime law, still effective till the late 16th century. 45
In 14th-century Majorca, the so-called taules de navegar (Lat. tabulas
navigandi) are none other than mappaemundi or navigational charts
which have been copied onto more than one piece of vellum… and pasted
onto wooden panels (often each chart onto two panels that could be folded
together to keep the map safely inside). 46 This kind of format, which
characterizes the Catalan Atlas, became popular among Majorcan
chartmakers of the period. In Venice, the influential Pietro Vesconte (fl.
1311-27) also named his atlases, mounted on wooden panels, as tabula(s),
so that he could distinguish them from his parchment charts, which he
43
Lo scrittore di queste righe ed autore di questa tavola (al-Ðabla) dice: Ciò è riferito
dall’autore del Nuzhat al-mushtāq...: cf. Nallino 1944, pp. 535-36, 540. See footnote 15.
44
See K. Lokotsch: Etymologisches Wörterbuch der europäischen (germanischen,
romanischen und Slavischen) Wörter orientalischen Ursprungs, Heidelberg, 1927, p. 156
(no. 1971). On the meaning of Ðabla as a drum, see H. Farmer: “¼andj,” EI2 9 (1997), p.
11; idem: “Æabl” and “Æabl-Khāna,” both in EI2 10 (2000). For other meanings:
Supplément, vol. 2, pp. 26-27; Lexicon, vol. 5, p. 1828; and M. A. Cherbonneau:
“Définition lexigraphique de plusieurs mots usités dans le langage de l’Afrique
septentrionale,” Journal Asiatique 13 (1849), (pp. 63-70, 537-51) 545 (planchette longue
et fendue à une de ses extrémités, dont se servent les brodeurs pour maintenir leur
ouvrage).
45
Other meanings in Medieval Latin are collected by Du Cange (C. du Fresne): Glossarium
mediae et infimae latinitatis (Paris, 1678). Reissued by L. Favre, Niort, 1883-87 (10 vols.),
vol. 8, pp. 4-8ff.; in which tabula portus / tabula maris refers to a mid 13th-century register
of coastal localities for tax purposes.
46
Cf. Pujades 2007, p. 109 (see also pp. 112-13).
242
M. Herrera-Casais
called carta(s). 47 According to this, the meaning of Ðabla / tabula suggests
a specific connection with the writing support on which the atlases were
made, either on wooden panels or paper paste folios, as in the case of
al-Sharafī’s atlases. The latter were bound in book format, as was often
the practice for 16th-century European and Ottoman atlases, for instance
Pīrī Re’īs’ Ba¬rīye (Book on sailing instructions, first version 1521 and
second version 1526). 48
9. Contents and composition
The core of al-Sharafī’s atlases comprises a set of sectional sea charts of
the Mediterranean and the Black Sea, which confers the nautical character
to the works. Only the second atlas incorporates a separate wind rose with
the names and directions of the thirty-two navigational winds. The charts
are introduced by a medieval cartography which describes a picture of the
universe in descending order from the celestial spheres to the Earth,
before focusing on the Mediterranean space. In the second larger atlas,
this consists of:
1. A cosmological scheme derived from al-Jaghmīnī’s treatise on the
fundaments of theoretical astronomy (Al-Mulakhkha½ fī l-hay’a
al-basīÐa, comp. 1221-22). 49 This was originally preceded by a scheme
of the phases of the moon which is now missing. Both elements are
omitted in the first atlas.
2. A scheme of Islamic sacred geography to be used for the determination
of the qibla or sacred direction towards the Kacba in Mecca. 50 This kind
47
Cf. Campbell 1987, p. 376. This terminology is used in Vesconte’s authorship inscriptions:
see Cartes nautiques sur vélin, p. 10 (1313 atlas); and Kretschmer 1909, pp. 111-12, 11617 (1313, 1318 and 1321 atlases).
48
Some European examples are described by Astengo 2007, pp. 183-85. See also Soucek’s
(1996, p. 108ff.) detailed codicological description of the late 17th-century Ottoman
Turkish atlas from the Khalili collection.
49
See the translation by G. Rudloff & A. Hochheim: “Die Astronomie des Ma¬mūd ibn
Mu¬ammed ibn cOmar al-Ğagmīnī,” Zeitschrift der Deutschen Morgenländischen
Gesellschaft 47 (1893), pp. 213-75; also S. P. Ragep: “al-Jaghmīnī,” in Th. Hockey et al.
(eds.): Biographical Encyclopedia of Astronomers, New York, 2007, vol. 2, pp. 584-85.
50
For sketches of the schemes, see Mappae arabicae, vol. 5, pp. 154-55 (atlas 1551) and one
of the guard leaves of the Atlas of Islam (atlas 1571). A colour picture of the scheme in the
The nautical atlases of cAlī al-Sharafī
243
of geography involves a notion of the Kacba as the centre of the world.
In both atlases, the scheme is inscribed on a wind rose which hints at
the adjustment of this distinctive nautical device to the service of
Islamic rituals in the Mediterranean.
3. A circular world map that al-Sharafī says to have drawn from al-Idrīsī’s
Geography, one of his chief authorities. 51 Only in the second atlas, this
is accompanied by a description of the shape of the Earth and the seas,
which is again an almost literal copy of the introduction to al-Idrīsī’s
text. 52 This information is summarized in the inscriptions of the sea
charts of the world by al-Sharafī and his son.
Further calendrical and folk astronomical material is inserted at the
beginning and the end of each atlas (before and after the cartography).
This includes a calendar of lunar mansions, a monthly almanac and data
on shadow lengths in the 1551 atlas, in which they appear in a smart
tabular presentation of didactic purposes. 53 To these, the 1571 version
first atlas is in The History of Cartography, vol. 2:1 (1992), pl. 13. D. King (World-Maps
for Finding the Direction and Distance to Mecca: Innovation and Tradition in Islamic
Science, London/Leiden, 1999) mentions both on pp. 52, 54, fig. 2.3.4; and in his article
“Makka (4. As the Centre of the World),” EI2 6 (1991). This author has published
extensively on folk and mathematical aspects of Islamic sacred geography, see idem (& R.
Lorch): “Qibla Charts, Qibla Maps, and Related Instruments,” in The History of
Cartography, vol. 2:1 (1992), pp. 189-205; idem: “The Sacred Geography of Islam.” In T.
Koetsier & L. Bergmans (eds.): Mathematics and the Divine: A Historical Study,
Amsterdam (etc.), 2005, pp. 161-78.
51
This appears at the beginning of some manuscripts of al-Idrīsī’s Geography, see Maqbul
1992, pp. 160ff., 173ff. (appendix), and figs. 7.1-7.5, 7.21, pl. 11. However, some scholars
claim that al-Idrīsī’s world map is derived from an archetype in the Book of Curiosities
(Kitāb Gharā’ib al-funūn wa-mula¬ al-cuyūn, comp. Egypt ca. 1050), which is only extant
in a 13th-century copy: cf. J. Johns & E. Savage: “The Book of Curiosities: A Newly
Discovered Series of Islamic Maps,” Imago Mundi 55 (2003), (pp. 7-24), 13-14, pl. 3.
52
See al-Idrīsī: Opus geographicum (Arabic edition), Naples/Rome, 1970-84, fasc. 1, pp.
3-14; and Géographie d’Edrisi. Translation by P. A. Jaubert, Paris, 1836-40, vol. 1, pp. 1-9
(revised by A. Nef & H. Bresc: Paris, 1999).
53
On tables as part of the visual language of science, see E. Savage: “Ibn Baklarish in the
Arabic Tradition of Synonymatic Texts and Tabular Presentations.” In Ch. Burnett (ed.):
Ibn Baklarish’s Book of Simples: Medical Remedies between Three Faiths in TwelfthCentury Spain, Oxford, 2008, pp. 113-32.
244
M. Herrera-Casais
adds other calendrical tables for finding the day of the week in which the
Arab and non-Arab (Julian) years begin (see below the contents of each
atlas). Contrary to the first one, all the additional material in this atlas is
presented in a functional format, with appended textual instructions. The
calendars of lunar mansions in both atlases are a distinctive element of the
Maghrebi nautical cartography. Three other such calendars appear in a
wheel shape in the charts of al-Æanjī and al-Mursī, as well as the Ottoman
Turkish one of ©ājj Abū l-©asan. These sources situate the lunar
mansions in a navigational context, but provide no definite evidence of
their practical use, perhaps as chronological indicators, by Muslim sea
captains of the Mediterranean. The choice of calendrical and astronomical
material that is also recorded in luxurious nautical atlases from Majorca
and Venice has not yet been fully surveyed. 54 This poses difficulties for
comparative research with al-Sharafī’s atlases. Of especial interest is the
Catalan Atlas (ca. 1375) which already shows a sumptuous cosmological
scheme with both the lunar mansions and the moon phases, besides outer
calendrical rings for the computation of the Julian year. 55
The similar contents and structure of al-Sharafī’s atlases suggests a
professional systematic method for their composition. Actually, he seems
to have worked with a set of cartographic models that could be combined
each time with different additional material at market demand. Such a
practice would have favoured a regular production of other comparable
atlases for a period possibly longer than the twenty years separating the
two extant examples.
54
See some remarks in Campbell 1987, pp. 446-48 and Pujades 2007, pp. 177-78.
55
See L’Atles Català (fol. II), with a study by J. Samsó in vol. 1, pp. 44-56 (quoted as Samsó
2005 in the final bibliography).
The nautical atlases of cAlī al-Sharafī
245
Table of contents of the 1551 atlas 56
Foliation
fol. 1v
fol. 2r
fol. 2v
fol. 3r
fol. 3v
fol. 4r
fol. 4v
fol. 5r
fol. 5v
fol. 6r
fol. 6v
fol. 7r
fol. 7v
fol. 8r
Contents
Front page with authorship and dating inscription
(T1) Calendrical table of lunar mansions
Scheme of sacred geography (oriented southwards)
Circular world map (oriented southwards)
(C1) Sea chart of the Iberian Peninsula and the western Maghreb
(oriented southwards)
(C2) Sea chart of the west central Mediterranean
(oriented southwards)
(C3) Sea chart of Italy and the Adriatic Sea (oriented southwards)
(C4) Sea chart of the Black Sea
(with the North black arrow mistakenly pointing eastwards)
(C5) Sea chart of the eastern Mediterranean (oriented northwards)
(C6) Sea chart of the Aegean Sea and the east central
Mediterranean (oriented northwards)
(C7) Sea chart of the central Maghreb with Sicily
(oriented northwards)
Circular table of shadow lengths
(T2) Monthly almanac (January – June)
(T2) Monthly almanac (July – December). Colophon
Table of contents of the 1571 atlas
Foliation
fol. 1r
fol. 1v
fol. 2r
fols. 2v-3r
fol. 3r
56
Contents
Front page with authorship and dating inscription
Instructions for T1 and T2
Instructions for the scheme of moon phases (fols. 1v–2r)
Description of the scheme of cosmology
Instructions for the scheme of sacred geography
Description of the nautical wind rose
Description of the shape of the Earth and the seas
Definition of sea chart of the world. First colophon
Abbreviations: T = table; C = sectional sea chart.
246
fol. 3v
missing
folio
fol. 4r
fol. 4v
fol. 5r
fol. 5v
fol. 6r
fol. 6v
fol. 7r
fol. 7v
fol. 8r
fol. 8v
fol. 9r
fol. 9v
fol. 10r
fol. 10v
fol. 11r
fol. 11v
fol. 12r
fol. 12v
fol. 13r
M. Herrera-Casais
(T1) First calendrical table for finding the weekday in which
the Arab months and years begin
[(T2) Second calendrical table of Arab months and years]
[Scheme of the phases of the moon]
Scheme of cosmology (oriented southwards)
Scheme of sacred geography (oriented southwards)
Nautical wind rose of 32 directions (oriented southwards)
Circular world map (oriented southwards)
(C1) Sea chart of the central Maghreb with Sicily
(oriented southwards)
(C2) Sea chart of Italy and the Adriatic Sea (oriented southwards)
(C3) Sea chart of the Iberian Peninsula and the western Maghreb
(oriented southwards)
(C4) Sea chart of the west central Mediterranean
(oriented southwards)
(C5) Sea chart of the Aegean Sea and the east central
Mediterranean (oriented southwards)
(C6) Sea chart of the Black Sea (oriented westwards)
(C7) Sea chart of the eastern Mediterranean
(oriented southwards)
(T3) Calendrical table of lunar mansions
(T4) Calendrical table for finding the weekday in which the
non-Arab (Julian) months and years begin
(T5) Table of shadow lengths
blank page
Instructions for T3 and T4 (fols. 11v–12r)
(T5) Instructions for T5
Description of the Mediterranean and the Atlantic
List of sources. Second colophon
The nautical atlases of cAlī al-Sharafī
247
III. THE SECTIONAL CHARTS
10. Geographical distribution and coastline layout
Al-Sharafī’s atlases offer an almost identical set of seven small sea charts
of the Mediterranean and the Black Sea. Most of them, especially in the
1551 atlas, are correlatively organized in facing pages, oriented
southwards. They focus on the profile of the coastline and, except for the
miniature of a pomegranate representing Granada, and a sketchy image of
Venice (this only in the second atlas), 57 the inland territories are left
empty. Six charts cover different sections of the Mediterranean, from the
Atlantic shores of the Iberian Peninsula and the Maghreb as far as the
Levant, and one chart of a smaller scale is dedicated to the Black Sea (see
their arrangement in the tables of contents of each atlas). The Iberian
Peninsula is divided between two charts, as the section of coastline
comprising Valencia and Catalonia (besides the Balearic Islands) is
included in the chart of the west central Mediterranean (figs. 3, 4). The
latter shows the Catalan coastline from Dénia (Dāniya*) to Cap
d’Aiguafreda (Qāb dī-Kūfrādā) in the first atlas, and to Barcelona
(Barshinūna*) in the second; in addition to the French and northwest
Italian coastlines as far as Piombino (Blubīn*).
The geographical distribution and coastline layout of the charts is
analogous in both atlases and probably derived from a common model or
master copy. The shape of mainland coastlines was transferred from this
model by means of a dry point technique which is well attested in
Mediterranean nautical cartography (for example, in Gabriel de Vallseca’s
1439 chart) 58 and described in detail in the Spanish nautical treatise of
Martín Cortés (1551). 59 All islands are drawn freehand. The profile of the
coastline in the 1571 atlas is carelessly, even exaggeratedly, distorted in
57
Venice is the most frequently depicted city in Mediterranean nautical cartography. Herrera
2008, p. 292 studies the emblem of Venice in al-Æanjī’s chart (compared with Albertin de
Virga’s 1409 chart), and Herrera 2009 in al-Mursī’s (compared with Bertrán & Ripoll’s
1456 chart). See Sáenz 2007, p. 453ff. for Majorcan and Italian examples.
58
I am grateful to Olga López (Museu Marítim, Barcelona) who allowed my access to the
original chart. See footnote 16.
59
See M. Cortés Albácar: Breve compendio de la esfera y del arte de navegar. Edition by M.
Cuesta Domingo, Madrid, 1990, pp. 217-19. The technique is discussed by Campbell
1987, p. 391 and Pujades 2007, p. 212ff.
248
M. Herrera-Casais
some regions. This is obvious in the Atlantic shore of the Iberian
Peninsula, extending from Cape Finisterre to the Bay of Cadiz and the
Straits of Gibraltar. Other problematic regions in the same atlas are the
Aegean seashores and the Peloponnese Peninsula, where prominent Cape
Matapan (named Māniya in the first atlas) is missing altogether. The main
Mediterranean islands, such as Sicily, Corsica and Sardinia, are also
visibly deformed.
11. Coastal place names
Apart from the more accurate coastline layout, the place name density is
also higher in the 1551 atlas, which provides almost seven hundred labels,
about a hundred more than the 1571 atlas. The significant toponymic
variants recorded between them can be explained by a different
combination of sources in each atlas, besides writing economy in the
second functional version. In order to accommodate a higher number of
place names, these were written in smaller size in the first atlas.
In all Maghrebi chartmaking, including al-Sharafī’s atlases, place
names were apparently inserted (and were to be read) starting in the
western Maghreb, to proceed around the Mediterranean coastline and end
up in the Iberian Peninsula or further north Atlantic, that is, in a counterclockwise direction when the charts are oriented northwards. This pattern
is equally observed in the work of ©ājj Abū l-©asan. The presentation of
the charts in Pīrī Re’īs’ Ottoman Turkish Ba¬rīye (1526) also maintains a
counter-clockwise arrangement around the Mediterranean, though starting
at the Dardanelles. 60 This distinctive feature might have originated in the
traditionally southward orientation of Islamic cartography, and is
coincidental with the Arabic geographies of al-Bakrī (d. 1094), al-Idrīsī
(fl. 1154) and Ibn Sacīd (1213-86) which in the Maghreb describe coastal
itineraries from West to East. 61 Indeed, the regional maps of al-Idrīsī’s
Geography were already presented in this order. 62 We find the opposite in
Majorcan and Venetian chartmaking where place names were to be read
60
Cf. Soucek 1992, p. 275. This is explained in Ba¬rīye, vol. 2, fol. 139r.
61
Some of these itineraries are examined by T. Lewicki: “Les voies maritimes de la
méditerranée dans le haut Moyen Age d’après les sources arabes,” in La navigazione
mediterranea nell’Alto Medioevo, Spoleto, 1978, vol. 2, pp. 439-80.
62
See footnote 11.
The nautical atlases of cAlī al-Sharafī
249
in a clockwise direction, starting north and ending south in the western
Maghreb (with the charts oriented northwards).
11.1. In the Islamic Mediterranean
The sequence of Arabic coastal place names in al-Sharafī’s atlases is
mostly attested in the 15th-century charts of al-Æanjī and al-Mursī, though
no direct connection can be established between them. They all offer an
exhaustive toponymic coverage of the Islamic Mediterranean which is one
of the main contributions of Maghrebi chartmakers to nautical
cartography. The compilation of such data probably benefited from the
earlier tradition of Arabic geography, but perhaps also from first hand
knowledge of the Islamic seafaring folklore in the Mediterranean. As we
have seen, al-Sharafī states to have used his father and grandfather’s
production, as well as al-Andalusī’s atlas (comp. before 1571).
Nonetheless, he might have had further knowledge of Maghrebi and
Andalusian chartmaking. No other Arabic cartographic source is
mentioned, except for the medieval maps of al-Idrīsī’s Geography.
Al-Sharafī’s collection of Arabic place names is less comprehensive
than that of his counterparts al-Æanjī and al-Mursī in all the Islamic
Mediterranean, that is, in North Africa, particularly in the seashores
between Tripoli and Alexandria, in addition to the Levant and the Iberian
Peninsula. Still, he shows a special concern for coastal hydrology,
meaning river mouths and water sources; for instance, in the western
Maghreb. He also provides many place names for minor localities
(including a few tribal settlements) and physical geography (primarily
capes and deltas) which are rarely attested, and possibly unknown, in
Majorcan and Venetian sea charts. In fact, Islamic regions habitually afar
from the maritime trading routes of European vessels, such as the Gulf of
Sirte and Marmarica, were accordingly furnished with lesser details in
their cartographies.
The Arabic place names in red ink for capital cities and major
geographical landforms are generally repeated in al-Æanjī and al-Mursī’s
charts, as well as al-Sharafī’s atlases. In North Africa, the latter records
almost fifty such names for the section of coastline between Ceuta and
Alexandria, which are the following: Sabta* (Ceuta), Targha*, Bādis*
(Rock of Vélez de la Gomera), Kha½ā½a*, Malīla* (Melilla), Mulūya*,
Hunayn* (not ©unayn), Wāhrān* (Oran), Mustaghānim*, Tanas*,
Brishk* (Biskra), Sharshāl* (Cherchell, in red ink only in the first atlas),
250
M. Herrera-Casais
al-Jazā’ir* (Algiers), Tadallīs* (Dellys), Bijāya* (Bougie), Jījil* (Jijelli),
al-Qull* (Collo), Ustūra* (Stora), Būna* (Annaba), Æabarqa* (Tabarca),
JāliÐa* (La Galite), Binzart* (Bizerta), Tūnis* (Tunis), Iqlībiya* (Kelibia),
jūn al-Madfūn* (in the Gulf of Hammamet, only in the second atlas),
Ihrīqliya* (Hergla), Sūsa* (Sousse), al-Mahdīya*, ¼afāqus* (Sfax),
Qarqana* (Kerkena), Qābis* (Gabes), Jarba* (Jerba), Zuwāgha* (in red
ink only in the first atlas), IÐrābulus* (Tripoli), Tājūra*, Labda* (Lebda),
Misrāta* (Misurata), ra’s Surt* (only in the second atlas), Namārish*,
al-BayÅā’* (in red ink only in the first atlas), Zināra*, Birnīq* (Benghazi),
Æulmayta* (Ptolemais), (ra’s) cAbdūn* (in red ink only in the second
atlas), ra’s Awtān* (Cape Sem), (ra’s) al-Hilāl*, ra’s al-Tīn*, Æubruq*
(Tobruk), Lukka*, ra’s al-Milā¬* (ra’s al-Mil¬), al-Æarqāwī* (in red ink
only in the first atlas), ra’s juzur al-©amām*, ra’s al-Kanā’is*,
Iskandarīya* (Alexandria).
Al-Sharafī maintains conventional Arabic place names in the Iberian
Peninsula, above all in Andalusia, which can be interpreted as a sign of
conservatism, and perhaps also as a political statement. There we still find
Burtuqāl* (Oporto), Æarf al-Gharb (Cape São Vicente), Ishbīliya*
(Seville), Jabal al-Fat¬* (Gibraltar, only in the first atlas), al-Jazīrat
al-KhaÅrā’* (Algeciras) or al-Munakkab* (Almuñécar), among others.
However, he transliterates the Catalan form of distinctive toponyms,
sometimes of Arabic origin, in the seashores of Valencia and Catalonia,
such as al-Bufīrā (l’Albufera, from Ar. al-bu¬ayra), qāb Larkūdrā (Cap
d’Alcodra, Cap de l’Horta), qāb di-Lalūyub (Cap de l’Aljup) and qāb
dī-Kūfrādā / dī-Kwafrādā (Cap d’Aiguafreda). 63 These examples,
probably copied from a Majorcan source, are unrecorded in the earlier
anonymous Maghreb Chart (ca. 1325-50). 64 In the Iberian Peninsula, the
latter provides a similar number of place names as al-Sharafī, but stays
closer to the Arabic tradition.
63
See R. Pujades: “La toponímia litoral del País Valencià en la cartografia portolana
medieval,” in Congrés internacional de toponímia i onomàstica catalanes (Valencia
2001), Valencia, 2002, (pp. 357-74) 363, 368, 369; and Rosselló 2004, pp. 311-12 (Cap de
l’Aljup). They are also listed by Kretschmer 1909, pp. 584-85, 588. See footnote 69.
64
Vernet 1962 offers a complete transliteration and identification of place names in the
Maghreb Chart (these are collected by Pinna 1996, vol. 2, p. 118ff.). See also the earlier
work by T. Fischer: “Die arabische Seekarte der Ambrosiana,” Sammlung mittelalterlicher
Welt- und Seekarten. Italienischen Ursprungs und aus italienischen Bibliotheken und
Archiven. Venice, 1886, pp. 219-45 (repr. Amsterdam, 1961).
The nautical atlases of cAlī al-Sharafī
251
As for the Aegean region (including the Balkan shore), which in the
16th century was exposed to Ottoman rule, al-Sharafī’s data is mainly
derived from a European (Majorcan) chart (see the comparative table
below). Here, significant variants are observed between al-Sharafī and Pīrī
Re’īs, whose Ba¬rīye logically favours Ottoman Turkish place names not
only in the Aegean, but also in Mediterranean Anatolia, even when
describing the location of Venetian and Genoese enclaves. 65 Nevertheless,
al-Sharafī incorporates al-Būghāz* which is the Ottoman Turkish
designation for the Dardanelles, also used by Pīrī Re’īs. 66 Another
example of Ottoman Turkish influence is Aynabakhtī* for Lepanto, which
is only included in the first atlas. If these were copied from the atlas of
al-Andalusī, who lived in Istanbul, the latter would have been al-Sharafī’s
source since 1551. As for the name of the Ottoman capital, al-Sharafī
records IsÐanbūl*, the location of which is highlighted with a prominent
flag. The first atlas adds the European designation Bīrā* (Pera) for nearby
Galata. Both IsÐanbūl* and Bīrā* (but not al-Būghāz) are attested in
al-Mursī’s chart, composed almost a decade after the Ottoman conquest of
the city (1453). Despite this historical fact, the name Constantinople
remained in vogue in Majorcan and Venetian nautical cartography long
after this date. 67
11.2. In Mediterranean Europe and the Black Sea
In al-Sharafī’s atlases, place names and geographical terminology for the
most part of Mediterranean Europe and the Black Sea are systematically
derived from a European, most probably Majorcan rather than Venetian,
source. Indeed, a detailed comparison of al-Sharafī’s sequence of coastal
place names with that of the more comprehensive Estense World Map
(mid-15th century) reveals frequent coincidence in both areas. This affinity
suggests influence from the Majorcan chart initially used by al-Sharafī’s
grandfather for his sea chart of the world. At this stage, it cannot be
determined whether al-Sharafī consulted this or another Majorcan model,
65
See Ba¬rīye, vols. 1-2. On Pīrī Re’īs’ sources, see Soucek 1992, pp. 277-79.
66
See Ba¬rīye, vol. 1, fol. 43v ff.; and V. Parry: “Boghaz-iči,” EI2 1 (1960); idem:
“Čana±-±ale Boghazi,” EI2 2 (1965), the latter with a reference to sea charts.
67
Cf. Campbell 1987, pp. 399, 400; some examples are given by Kretschmer 1909, pp. 64041. On the name Istanbul, see H. İnalcık: “Istanbul,” EI2 4 (1978), p. 224.
252
M. Herrera-Casais
presumably kept at the family workshop, or relied directly on his
grandfather’s work. As the extant sea charts of the world by al-Sharafī and
his son are dependent on the grandfather’s prototype, a comparative study
of their toponymy with that of al-Sharafī’s atlases would help to estimate
the impact of the Majorcan chart on them. The combination of Arabic and
European sources is already attested in the Maghreb Chart, al-Æanjī and
al-Mursī. 68 As such, it is likely that al-Andalusī’s atlas had also played an
intermediary role in the transmission of European toponymy to al-Sharafī.
It is worth to note that al-Sharafī records several times Sant Yūrdī,
which is a transliteration of the Catalan Sant Jordi, rather than the Italian
San Giorgio which could have produced San Jurjīs or San Jurj. Among
other examples, it stands for Sveti Juraj (Jurjevo) in the Adriatic, Cape
Agios Georgios in the Aegean (this only in the first atlas), and Sfantu
Gheorgue next to the Danube Delta. The first atlas adds Sant alone for
Sant Jordi d’Alfama in Catalonia, which is already given in the Maghreb
Chart (SānÐ Jūrj). Both atlases include Artadur* (Cape Artatur) which is a
typically Majorcan place name in the Adriatic. 69 Two hispanicisms are
noticeable: Insulā d-Ālbī (Rab Island, again in the Adriatic), in the first
atlas, and Insulārūsā (Dzharylgach Island, in Crimea), in both atlases. The
latter is written in red ink only in the first one, in which it appears next to
the Arabic term al-qanÐara* (spit of land), for the Perekop Isthmus. Both
hispanicisms are recorded in the Catalan Atlas (Albi, ínsula Rossa) and
the Estense World Map (insula de alb, insularosa). 70 In addition,
al-Sharafī renders the Italian Venice into Arabic as Finīziya wa-hiya lBunduqīya* (Venice, that is, al-Bunduqīya).
As for the geographical terminology, al-Sharafī uses the form qāb more
often than qābū, denoting a cape (Ar. ra’s, Ðarf). These are transliterations
of cap and cauo (cavo) respectively, which are familiar variants to both
Majorcan and Venetian chartmakers. However, the form qāb (cap) is
strongly associated to the Catalan language. 71
68
See Herrera 2008, pp. 296-98 (al-Æanjī); and footnote 64 (Maghreb Chart).
69
Cf. Campbell 1987, p. 425; and Pujades’ contribution to L’Atles Català, vol. 1, p. 42. In
addition, Pujades 2007, p. 349ff. (appendix) lists the toponymy of the Adriatic and
Catalonian seashores in 14th-15th-century Majorcan and Venetian charts.
70
See L’Atles Català (fol. IV), vol. 1, pp. 133, 139; and MCE, pp. 137, 160.
71
For further discussion, see Rosselló 2000, pp. 83-84; and Rosselló 2004, pp. 321-24.
The nautical atlases of cAlī al-Sharafī
253
12. The wind network
The charts in al-Sharafī’s 1551 atlas lack a wind network proper, as the
thirty-two wind directions radiate from the centre of the pages directly
towards the margins. The winds are drawn in the same way in the 1563
atlas by Jaume Olives. 72 In al-Sharafī, North is indicated with a black
arrow, usually placed at the bottom of the page. This North arrow already
appears in the Isolario of Bartolommeo dalli Sonetti’s (nickname of
Bartolommeo Zamberti), printed in Venice in 1485, 73 and is also a
distinctive feature of the charts in Pīrī Re’īs’ Ba¬rīye.
Al-Sharafī’s 1571 atlas shows a regular network of sixteen winds, with
decorative medallions in the secondary centres. The perimeter defined by
these centres, each marked with a pinhole, encloses the mapped surface in
circumference of about 19 cm diameter. It is unclear whether the wind
network was laid down earlier than the shape of the coastline. 74 In any
event, my close examination of the atlas under the microscope convinced
me that place names were inserted after both the network and the coastline
had been drawn.
In each atlas, the centre of the wind system is marked with a pinhole
deep enough to perforate the paper paste folios, so that one pinhole is
shared by the two charts on both sides of the folio. This pinhole was
probably made with a compass tip. The central point falls on identical
cartographic locations in both atlases, for instance: next to Almuñécar, in
the chart of the Iberian Peninsula; on the northern coast of Minorca, in the
chart of the west central Mediterranean; to the south of Ancona, in the
chart of the Adriatic; and in the Gulf of Argolis, in the chart showing the
Peloponnese (east central Mediterranean). This indicates that the centre of
the wind system was used as a reference point for drawing the coastline
around it. Such a technique would have also helped to imitate the master
72
This is kept at Olomouc (Czech Republic), Research Library: M II 33. The electronic
version is available at the Library website: < http://dig.vkol.cz/ >.
73
See G. Tolias: “Isolarii, Fifteenth to Seventeenth Century,” in The History of Cartography,
vol. 3:1 (2007), (pp. 263-84) 268-69 (Sonetti’s Isolario), 269-70 (Ba¬rīye). The isolarii
(books on islands) had a strong influence on the making of Ba¬rīye: see Soucek 1992, pp.
277-78; Soucek 1996, p. 20ff.
74
It was actually settled before the coastline in four 14th-15th-century charts from the British
Library: cf. Campbell 1987, p. 390. This topic is discussed by Pujades 2007, p. 188ff. See
footnote 59.
254
M. Herrera-Casais
copy proportionally, and this supports the idea that the charts of both
atlases are derived from a common model.
13. The pomegranate
A rare feature of al-Sharafī’s cartography is the miniature of a
pomegranate placed on the Iberian Peninsula, to be interpreted as an icon
for the Nasrid kingdom of Granada. 75 The pomegranate appears on both
atlases as well as the sea charts of the world by al-Sharafī and his son.
This icon fuels an evocative nostalgic image of al-Andalus, and implies
moral support to emigrant moriscos from the Iberian Peninsula in the
context of the Ottoman defence of Mediterranean Islam. It is uncertain
whether the insertion of the pomegranate was an original idea of
al-Sharafī, or it was copied from the earlier cartography of his family, the
atlas of A¬mad al-Andalusī, who maybe was a morisco himself, or other
sources. He could have also learned about its symbolism from the many
moriscos that had settled in Tunisia and kept alive their Andalusian
traditions. 76 In any event, the fruit appears already in Jaume Bertrán’s
1489 chart where a figure, perhaps King Ferdinand the Catholic, holds a
pomegranate in his hand anticipating the fall of Islamic Granada (1492). 77
Al-Sharafī draws the rivers Guadalquivir and Segura flowing from a
common source at the top of the pomegranate. This source is distinctively
placed at the top of Mount Segura by Majorcan chartmakers, whose
depiction of the river courses towards their mouths in the Atlantic and the
Mediterranean can be traced back to al-Idrīsī’s Geography. 78 ( 79 )
75
This is studied by Herrera 2009.
76
See Epalza 1992, pp. 261-76; and L. Bernabé: “Notas sobre la cohesión de la comunidad
morisca más allá de su expulsión de España,” Al-QanÐara 29:2 (2008), pp. 307-32.
77
This chart is mentioned by Sáenz 2008, pp. 631-32; idem: “La Reconquista cartográfica. El
Islam peninsular en la cartografía medieval hispana,” Treballs de la Societat Catalana de
Geografia 61-62 (2006), (pp. 279-301) fig. 8.
78
See Herrera 2009. The image of Mount Segura in Majorcan chartmaking is described by
Rey & García 1960, p. 29; and Sáenz 2008, p. 369.
79
This table only enters toponyms recorded in al-Sharafī’s atlases, though L’Atles Català
(fol. IV, in vol. 1, pp. 139, 141) and the Estense World Map (MCE, pp. 165, 167, 168-69)
offer more details for the same section of coastline. See also Kretschmer 1909, pp. 639,
652-53 for other European charts.
80
MCE
(ca. 1450)
al-Sharafī
Atlas 1551
al-Sharafī
Atlas 1571
Place-name identification
Sesmire
laxmiro*
Izmīra*
Izmīr(a)*
İzmir
Foya
foya*
Fūja*
Fūja*
Yenifoça
Cristo (golf de Laler)
cristo
Krīstū
Krīstū
at Çandarlı Körfezi
Stingary
stringani
Stinkām
Stinkām
(cape)
San Zorzo
santiordi
Sant Yūrdī
Sant Yūrdī
next to Ayvalık
Landermiti*
llaudermeti*
Ladrimītī*
Ladrimītī*
Edremit
cauo Sancta-Maria
cauo sta maria
Qāb Sant Māriyā
Qāb
Baba Burnu
Dardanelo
Dardanello
al-Būghāz*
al-Būghāz*
Çanakkale Boğazı
(Eng. Dardanelles)
Gallipolli*
Gallipoli*
Kālibūlī*
Kālibūlī*
Gelibolu (Eng. Gallipoli)
G. de cardia
Kardiyā
Fīlā
80
The nautical atlases of cAlī al-Sharafī
Catalan Atlas
(ca. 1375)
( 80 )
Saros Körfezi
255
This already appears as Golfo di Cristo in Vesconte’s works: cf. Kretschmer 1909, p. 653.
Comparative table of place names in north-western Anatolia
and the Dardanelles79
256
M. Herrera-Casais
Fig. 3: AL-SHARAFĪ’S 1551 ATLAS
Sectional sea chart of the Iberian Peninsula and the western Maghreb (fol. 3v),
oriented southwards in the original (northwards in the picture).
Courtesy of the Bibliothèque nationale de France, Paris (MS arabe 2278).
The nautical atlases of cAlī al-Sharafī
Fig. 4: AL-SHARAFĪ’S 1571 ATLAS
Sectional sea chart of the Iberian Peninsula and the western Maghreb (fol. 7r),
oriented southwards in the original (northwards in the picture).
Courtesy of the Bodleian Library, Oxford (MS Marsh 294).
257
258
M. Herrera-Casais
14. Conclusion
The two extant atlases of the 16th-century Tunisian chartmaker and scholar
c
Alī al-Sharafī are indispensable documents for understanding the history
and development of Maghrebi nautical cartography in the Mediterranean
context. They provide precious information on al-Sharafī’s chartmaking
techniques, production and sources, which include the output of his own
family. Moreover, their study raises interesting questions on the working
conditions in which Maghrebi and Andalusian chartmakers flourished and
the impact of their legacy. The emergence in Istanbul of a certain A¬mad
al-Andalusī, which is quoted as one of al-Sharafī’s sources, also brings
forward the issue of the Andalusian involvement in Mediterranean
nautical cartography.
The similar contents and structure of al-Sharafī’s atlases imply a
systematic method of composition which would have favoured a regular
production of such works at market demand. This might have allowed
al-Sharafī to earn a living as a professional chartmaker for about three
decades. Furthermore, the second atlas informs us of three otherwise
unknown sea charts of the world which he had probably made before
1571. A total of six charts of this kind, four of them by al-Sharafī himself,
can now be safely attributed to the Sharafī family. Only two of these, one
by al-Sharafī (1579) and another by his son, have survived.
The 1551 atlas is a visually striking work of art, whereas the one from
1571 is an extended functional version with a larger set of calendrical and
folk astronomical material. The latter atlas has been heavily used and is
worn out, in contrast to the first one which is preserved in excellent
condition. Both offer almost identical sea charts of the Mediterranean and
the Black Sea, which seem to have been derived from a common model or
master copy. They show the same geographical distribution, coastline
layout and drawing technique, though they are more carefully copied and
have a higher density of place names in the first atlas. Only the second
one supplies the charts with a network of wind directions and adds a
nautical wind rose.
The diverse origins of al-Sharafī’s sources provide good testimony to
the transfer of knowledge within the Maghrebi and Andalusian
chartmaking tradition, and to the positive influence of the Majorcan one
into the latter. Most of his Arabic coastal place names for the Islamic
Mediterranean are already attested in the 15th-century works of al-Æanjī
and al-Mursī, though no direct connection can be established between
The nautical atlases of cAlī al-Sharafī
259
them. They all offer exhaustive compilations of Arabic toponymy which
are an outstanding contribution of Maghrebi chartmakers to
Mediterranean cartography. On the contrary, al-Sharafī’s place names for
Mediterranean Europe and the Black Sea are mostly transliterated from a
15th or 16th-century Majorcan chart. The latter might have influenced alSharafī through his grandfather’s sea chart of the world for which a
Majorcan model was consulted. Other sources, such as al-Andalusī’s atlas,
might have played an intermediary role in the transmission of European,
and possibly Ottoman Turkish, toponymy to al-Sharafī. Political changes
in the Mediterranean also explain the inclusion of distinctive Majorcan
place names next to the traditional Arabic ones in the Iberian Peninsula, as
well as updated Ottoman Turkish designations in the Aegean and the
Balkans.
The combination of Islamic, mainly Arabic, and European sources,
manifest in all the extant Maghrebi nautical cartography, was encouraged
by the circulation of sea charts across the Mediterranean, within the frame
of migrations, trading interaction and cultural exchange. Indeed, the
Majorcan elements observed in the works of al-Æanjī and the Sharafīs
suggests the availability of Majorcan charts in Tunisia (at least in Tunis
and Sfax) between the 15th and 16th centuries. Further research should try
to evaluate the impact and spreading of Maghrebi chartmaking in
Majorcan and Venetian workshops of the same period.
15. Acknowledgements
This paper is based on my forthcoming doctoral dissertation Cartografía
náutica árabe en el siglo XVI: Los atlas de cAlī al-Šarafī de Sfax. I have
been able to consult the manuscripts of al-Sharafī’s atlases thanks to two
visits to the Bibliothèque nationale de France and the Bodleian Library.
These trips were generously supported by the research projects La
actividad científica de herencia árabe en la España del siglo XV: El caso
del Sexagenarium (Gobierno de Canarias PI2000/133) at the University of
La Laguna, and Cartografía náutica árabe en el contexto mediterráneo,
ca. 1300–1600: Influencias entre oriente y occidente (Ministerio de
Educación HUM2005–03375/FILO) at the University of Barcelona. I am
grateful to Mercè Comes (Barcelona) and Mohsen Zakeri (Frankfurt) for
their critical readings and useful suggestions, and to Colin Wakefield for
his kind help during my stay at the Bodleian (Oriental Collections).
260
M. Herrera-Casais
16. Bibliography 81
Astengo 2007 = Astengo, Corradino: “The Renaissance Chart Tradition
in the Mediterranean,” in The History of Cartography, vol. 3:1 (2007), pp.
174-262
Atlas of Islam = An Historical Atlas of Islam. First edition by W. C.
Brice, 1981; second edition by H. Kennedy, 2001. Leiden: Brill.
Ba¬rīye = Pīrī Re’īs: Kitāb-ı Ba¬rīye [1526]. Coord. by Ertuğrul Z.
Ökte. Facsimile of MS Istanbul, Süleymaniye Kütüphanesi: Ayasofya
2612. Edition and translation by V. Çabuk, T. Duran and R. Bragner.
Ankara: Ministry of Culture and Tourism of the Turkish Republic (etc.),
1988, 4 vols.
Cairo Survey = King, D. A.: A Survey of the Scientific Manuscripts in
the Egyptian National Library. Winona Lake: Eisenbrauns, 1986.
Campbell 1987 = Campbell, Tony: “Portolan Charts from the Late
Thirteenth Century to 1500,” in The History of Cartography, vol. 1
(1987), pp. 371-463.
Cartes nautiques = Foncin, Myriem – Destombes, Marcel & La
Roncière, Monique de: Catalogue des cartes nautiques sur vélin
conservées au Département des Cartes et Plans. Paris: Bibliothèque
nationale [de France], 1963.
Comes 2004 = Comes Maymó, Mercè: “La cartografia a Mallorca i a
Barcelona.” In J. Vernet and R. Parés (eds.): La ciència en la història dels
Països Catalans, vol. 1: Dels àrabs al Renaixement. Barcelona: Institut
d’Estudis Catalans and Valencia: Universitat de València, 2004, pp. 51573.
EI2 = The Encyclopaedia of Islam. Leiden: Brill, 1960-2004, 12 vols.
Epalza 1992 = Epalza Ferrer, Mikel de: Los moriscos antes y después
de la expulsión. Madrid: Mapfre, 1992.
GAL = Brockelmann, Carl: Geschichte der arabischen Litteratur.
Leiden: Brill, 1943-49 (2nd edition), 2 vols.; and Supplementbände.
Leiden: Brill, 1937-42, 3 vols.
Herrera 2008 = Herrera Casais, Mónica: “The 1413–14 Sea Chart of
A¬mad al-Æanjī.” In E. Calvo, M. Comes, R. Puig & M. Rius (eds.):
A Shared Legacy: Islamic Science East and West. Homage to Professor J.
M. Millàs Vallicrosa. Barcelona: Universitat de Barcelona, 2008, pp.
283-307.
81
Full references given in the footnotes are not repeated in the bibliography.
The nautical atlases of cAlī al-Sharafī
261
___ 2009 = “Granada en la cartografía náutica árabe, e identificación
de un modelo mallorquín para al-Mursī.” Al-QanÐara 30:2 (2009),
forthcoming.
Itinéraire du savoir en Tunisie = Annabi, Hassen – Chapoutot-Remadi,
Mounira & Kamarti, Samia (eds.): Itinéraire du savoir en Tunisie. Les
temps forts de l’histoire tunisienne. Paris: Institut du Monde Arabe and
Tunis: ALIF, 1995.
Kretschmer 1909 = Kretschmer, Konrad: Die italienischen Portolane
des Mittelalters. Ein Beitrag zur Geschichte der Kartographie und Nautik.
Berlin, 1909.
L’Atles Català = L’Atles Català 1375. El món i els dies. Facsimile of
MS Paris, Bibliothèque nationale de France: Espagnol 30. Studies by
Gabriel Llompart, Ramon J. Pujades, Julio Samsó. Barcelona:
Enciclopèdia Catalana, 2005, 2 vols.
Lexicon = Lane, Edward W.: An Arabic-English Lexicon.
London/Edinburgh: Williams & Norgate, 1863-93, 8 vols. (repr. Beirut:
Librairie du Liban, 1968).
Ma¬fūÞ 1982-86 = Ma¬fūÞ, Mu¬ammad: Tarājim al-mu’allifīn
al-tūnisīyīn. Beirut: Dār al-Gharb al-Islāmī, 1982-86, 5 vols.
Mappae arabicae = Miller, Konrad: Mappae arabicae: Arabische Weltund Länderkarten des 9.-13. Jahrhunderts. Stuttgart, 1926-31, 6 vols.
Maqbul 1992 = Maqbul, Ahmad S.: “Cartography of al-Sharīf
al-Idrīsī,” in The History of Cartography, vol. 2:1 (1992), pp. 156-74.
Maqdīsh, Nuzha = Maqdīsh, Ma¬mūd: Nuzhat al-anÞār fī cajā’ib
al-tawārīkh wa-l-akhbār. Edited by cAlī al-Zuwārī & Mu¬ammad
Ma¬fūÞ. Beirut: Dār al-Gharb al-Islāmī, 1988, 2 vols.
MCE = Il Mappamondo Catalano Estense. Facsimile dell’originale
conservato presso la Biblioteca Estense di Modena. Introduction by
Ernesto Milano. Transliteration and identification of inscriptions and
place names by Annalisa Battini. Dietikon, Zurich: Urs Graf
(Italian/German version), 1995.
MGC = Sezgin, Fuat: Mathematical Geography and Cartography in
Islam and Their Continuation in the Occident. Frankfurt a. M.: Institut für
Geschichte der Arabisch-Islamischen Wissenschaften, 2000-07, 4 vols.
(being vols. 10-13 of Geschichte des arabischen Schrifttums; vol. 13 only
in German).
Nallino 1944 = Nallino, Carlo A.: “Un mappamondo arabo disegnato
nel 1579 da cAlī ibn A¬mad ash-Sharafī di Sfax.” Raccolta di scritti editi
262
M. Herrera-Casais
e inediti a cura di Maria Nallino. Roma: Istituto per l’Oriente, vol. 5
(1944), pp. 533-48.
Nicoll 1821 = Nicoll, Alexander: Bibliothecae Bodleianae codicum
manuscriptorum orientalium catalogi. Partis secundae volumen primum
Arabicos complectens. Oxford (Oxonii, e Typographeo Clarendoniano),
1821.
Pinna 1996 = Pinna, Margherita: Il Mediterraneo e la Sardegna nella
cartografia musulmana (dall’VIII al XVI secolo). Nuoro: Istituto
Superiore Regionale Etnografico, [1996], 2 vols.
Pujades 2007 = Pujades i Bataller, Ramon J.: Les cartes portolanes. La
representació medieval d’una mar solcada. Barcelona: Lunwerg, 2007.
Rey & García 1960 = Rey Pastor, Julio & García Camarero, Ernesto:
La cartografía mallorquina. Madrid: Consejo Superior de Investigaciones
Científicas, 1960.
Rosselló 1995 = Rosselló i Verger, Vicenç M.: “Cartes i atles portolans
de les col·leccions espanyoles.” Portolans procedents de col·leccions
espanyoles: Segles XV-XVII [with catalogue references by Carme
Montaner]. Barcelona: Institut Cartogràfic de Catalunya, 1995, pp. 9-59.
___ 2000 = “Les cartes portolanes mallorquines.” In Cicle de
conferències sobre Història de la Cartografia: 10è curs La Cartografia
Catalana (22-26 de febrer de 1999). Barcelona: Institut Cartogràfic de
Catalunya, 2000, pp. 18-115.
___ 2004 = Toponímia, geografia i cartografia. Collected studies
edited by E. Casanova. Valencia: Universitat de València, 2004.
Sáenz 2007 = Sáenz-López, Sandra: Imagen y conocimiento del mundo
en la edad media a través de la cartografía hispana. Doctoral dissertation
presented at Universidad Complutense, Madrid, 2007, 2 vols.
Samsó 2005 = Samsó, Julio: “Una concepció de l’Univers.” In L’Atles
Català, vol. 1, pp. 44-56. [First version with Joan Casanovas:
“Cosmografia, astrologia i calendario.” In L’Atlas Català de Cresques
Abraham. Primera edició en el sis-cents aniversari de la seva realització,
1375-1975. Barcelona: Diáfora, 1975, pp. 23-36].
Seyā¬atnāme = Čelebi, Evliyā: Narrative of Travels in Europe, Asia
and Africa in the Seventeenth Century. Translation by J. von Hammer.
London: The Oriental Translation Fund of Great Britain and Ireland,
1834-46, 2 vols.
Soucek 1992 = Soucek, Svat: “Islamic Charting in the Mediterranean,”
in The History of Cartography, vol. 2:1, pp. 263-92.
The nautical atlases of cAlī al-Sharafī
263
Soucek 1996 = Piri Reis and Turkish Mapmaking after Columbus:
The Khalili Portolan Atlas. London (etc.): The Nour Foundation, 1996
(1st edition 1992).
Supplément = Dozy, Reinhart: Supplément aux dictionnaires arabes.
Leiden: Brill and Paris: Maisonneuve, 1927 (2nd ed.), 2 vols.
The History of Cartography (1987 & 1992) = Harley, J. Brian &
Woodward, David (eds.): The History of Cartography, vol. 1 (1987):
Cartography in Prehistoric, Ancient, and Medieval Europe and the
Mediterranean; and vol. 2:1 (1992): Cartography in the Traditional
Islamic and South Asian Societies. Chicago/London: The University of
Chicago Press.
The History of Cartography (2007) = Woodward, David (ed.): The
History of Cartography, vols. 3: 1-2 (2007): Cartography in the European
Renaissance. Chicago/London: The University of Chicago Press.
Vernet 1962 = Vernet Ginés, Joan: “The Maghreb Chart in the
Biblioteca Ambrosiana.” Imago Mundi 16 (1962), pp. 1-16.