The Transformation of al-Idrīsī’s Nuzhat al-Mushtāq in the Atlases of ʿAlī al-Sharafī of Sfax

Abstract

This article analyzes the ways in which ‘Alī al-Sharafī of Sfax, an atlas author of the sixteenth century, adapted excerpts from the work of the twelfth-century geographer Muḥammad al-Idrīsī as well as features of medieval European portolan charts. The article pursues a translation studies approach. It argues that, through the adaptation of both textual and graphic elements from Idrīsī’s work and the incorporation of other graphic and cartographic elements, al-Sharafī was in fact engaging in a practice of cultural translation which resulted in a unique and innovative textual and cartographic oeuvre. As such, al-Sharafī should be regarded not simply as a passive copyist of textual and cartographic elements, but indeed as an author who engaged in the creative re-signification of the traditions and practices that informed his work.

The question of authorship in translation is one that requires careful examination. While the source or ‘original’ work is always at the center of any translation endeavor and the voice of the author should always be retained, it is also true that the translator is in many ways a second author whose voice inevitably resonates in the final work. In this sense, the authorship of a translation is multilayered and complex. This applies to an even greater extent to less conventional types of translation, such as cultural translation between different media. The case at the heart of this study is an example of the latter. The focus is on the adaptation of a medieval geographical text by the twelfth-century Arab geographer Muḥammad al-Idrīsī (d. 1165) into maps and atlases by the sixteenth-century North African mapmaker ʿAlī b. Muḥammad al-Sharafī of Sfax (d. after 1579). Here we understand translation not as the transfer of a text from one language into another, but primarily as a transfer from a predominantly textual medium into a multi-media (textual, cartographic, and iconographic) product, supplemented by occasional transfers of other iconographic elements. At the same time it also represents a conscious modification of al-Idrīsī’s source text in the form of abridgements and adaptations in order to tailor it to the needs of the medium as well as to assert al-Sharafī’s claim to be the co-author of the work. In many genres of Arabic literature, an authorial persona is clearly identified in forewords, afterwords, and colophons.¹ Although this is also the case in al-Sharafī’s works, none of his colophons follows the classical model precisely—in terms of number, structure, or content—of Arabic and Islamic pre-modern works; but neither does it follow the model and structure of European atlases and charts.²

In al-Sharafī’s work, as we will see in more detail, the authorial voice becomes more pronounced through the development of a style and through particular choices made at various points in the body of each work. For instance, while (with one exception) he does not include the characteristic “chain of transmission” (isnād)³ of the authorities who supplied the knowledge required for the construction of his maps, he does repeatedly quote authorities in the fields of geography, cartography, astronomy, and Sufism from different epochs and places.⁴ In this way, he constructs a strategy of knowledge validation and inserts himself⁵ as one more authorial link in the historical chain of transmission of this body of multidisciplinary knowledge through time and space. He thus creates a pattern of multiple authorial voices, which is also common in Classical Arabic literature. Idrīsī’s Nuzhat al-mushtāq fī ikhtirāq al-āfāq (The Book of Pleasant Journeys into Faraway Lands, hereafter referred to simply as the Nuzha) is also an example of this. In it Idrīsī cites a number of geographical authorities, both Arabic and European, such as Ptolemy (d. c. 170) and al-Masʿūdī (d. 955).⁶ Lale Behzadi has rightly compared Idrīsī’s approach in this work to the concept of “polyphony”, in the sense that the author often steps back and allows other voices to be heard.⁷ Al-Sharafī’s works are likewise polyphonic in that they retain al-Idrīsī’s voice (among other sources), while also allowing his own authorial voice to be heard. His practice of repeatedly citing other sources should therefore be seen not as a sign of passivity or a lack of originality, but rather as a way of using other authorities to establish an authorial voice, in a quest to anchor a work within a broader tradition and to assert his authoritative status.⁸

This paper will thus focus on al-Sharafī’s decision to present himself as an author. We will explore this at three different levels: 1) at the material level, in terms of the transfer of geographical knowledge and cartographic practices from different sources and traditions into the format of a portolan atlas, the model adhered to by al-Sharafī in the composition of his atlases; 2) at the textual level, in terms of the selective adaptation of sources like al-Idrīsī’s work in a quest to create a new work, which would be sufficiently different to be considered an original product, while retaining enough features from its sources to render them recognizable; and 3) in terms of anchoring of the work in question within its local (Mediterranean) setting, through the selection of an appropriate lexical and symbolic repertoire.⁹

10.1 The Concept of Politics of Translation as an Approach to Analyzing al-Sharafī’s Works

Most studies dealing with the concept of politics of translation tend to focus on two overarching themes: the first relates to the sum of the contextual factors surrounding any given translation endeavor. These of course include questions of patronage and commissioning, but also broader issues related to transmission of knowledge across cultures, ranging from missionary projects to cultural propaganda, to name just two obvious examples.¹⁰ The second theme relates to the more technical aspects of translation, including lexical and rhetorical choices, which are informed by the cultural (and political) frameworks from which translation projects emerge. Examples of complex and new approaches to this notion abound, from Gayatri Spivak’s studies on feminist translation as solidarity with neglected voices in postcolonial contexts, to Lori Anne Ferrell’s study of stylistic choices in English Bible translations, or Kenneth Lloyd-Jones’s study of Erasmus’s critique of Ciceronianism in the early modern period.¹¹ Of special interest and useful for our analysis is the inclusion of the spatial dimension and cartography in translation studies. In his studies on the poetry of Ludovico Ariosto (d. 1533), Francesco Italiano referred to this as “the translation of geographies” or the carticity of a text.¹² The latter term, as understood by Robert Stockhammer, refers to a kind of writing that uses words to mirror the spatial characteristics of a map. According to Stockhammer, carticity of literary description exhibits an “affinity or distance to cartographical process or representation, not only with respect to the setting, but also […] with respect to the creatures to be located there”.¹³ This can be achieved through a detailed description of the topographic elements of the map, through a sequence that describes an itinerary, or by any other means through which the visual representation becomes indispensable for understanding the text, in a way reminiscent of a map.¹⁴ Related to these approaches, albeit comprising a somehow different aspect of politics of translation, is the question of how translators have used their newly conceived texts as legitimizing tools to signal their belonging to a given social or scholarly community.

In our study of al-Sharafī’s works, we look beyond conventional understandings of translation—that is, translation as a source-to-target language transfer—in order to consider other processes, such as dynamic equivalence, intersemiotic translation (recently called transmediation), and carticity.¹⁵ In this broader understanding of cartography as translation—indeed as a complex process of cultural negotiation—¹⁶ politics of translation should be linked to the positionality of the creator of such anthologies, textual collages, and compendiums. In other words, the decisions made in the process of assembling, blending, interpreting, and modifying sources in order to create a new product should be understood as modes of intellectual “social signaling” by any given translator or intellectual authority.

As a case study, we examine the appropriation of the above-mentioned Nuzha of al-Idrīsī in the works of ʿAlī al-Sharafī.¹⁷ By unravelling this practice of abridgement, adaptation, and reformulation, we argue that al-Sharafī was not engaged in a mere act of imitation through copying, but in one of active appropriation for the purpose of generating an original textual and visual product of his own. Through his use of al-Idrīsī’s text, al-Sharafī created a framework which allowed him to invest his work with an intellectually authoritative status and to position himself as an authorial persona. As we will see later, al-Sharafī mobilized cultural capital through references and praise for al-Idrīsī’s work, which in his time was a renowned classic of geographical knowledge and mapmaking. This enabled him to present himself as an heir to a long-established tradition of knowledge and mapmaking practiced in multiple Islamic circles. But these references were not his only modes of speaking with and through al-Idrīsī’s voice. Nor was the use of this work his only strategy for translating Majorcan and Italian sea charts and atlases into the knowledge contexts of the two urban societies in which he lived—Sfax and Qayrawan. Numerous other—textual as well as visual—features were used to modify the maps and their contextual elements, which he sought to translate into his two atlases. Creatively appropriating well-known texts, architectural decorations, calligraphies, paintings, materials, layouts, and metric formats were the major ways in which al-Sharafī innovatively transformed foreign charts and atlases and combined them with graphic elements that were more common in his cultural milieu. Hence, in our analysis of al-Sharafī’s work we understand politics of translation as a series of rhetorical and visual choices made by the translator—the mapmaker—to anchor his product within a broader intellectual tradition of textual and visual geographical knowledge in Arabic, and also to ground it in the Mediterranean context in which he lived.

10.2 ‘Alī al-Sharafī: Copyist or Author?

The work of ʿAlī al-Sharafī of Sfax, an otherwise little-known character associated with the production of two atlases and a grand-scale world map,¹⁸ challenges our understanding of the relationship between map production and appropriation and the reinterpretation of geographical knowledge in many ways. The dearth of information about him and about the context in which he worked makes it impossible to know, for example, whether being a mapmaker was his primary occupation or “professional identity”. We know that members of his family, in both previous and subsequent generations, were involved in mapmaking and astronomy,¹⁹ but it is not entirely clear what the mechanisms for transmitting the skills and knowledge associated with this activity were, nor in what kind of institutional setting this occurred.²⁰ Neither is it clear what kind of intellectual upbringing al-Sharafī had.²¹ Almost all the evidence we have derives directly from his works. Only they provide us with any clues about what mapmaking as an intellectual enterprise meant for him. As such, the graphic and textual decisions in his atlases (and in his world map, for that matter) give us some insight into his strategies of intellectual self-fashioning, which reveal the different layers of his authorial persona.

Al-Sharafī relied on already existing geographical knowledge to produce his works, drawing from it in the form of adapted quotations. In contrast to previous evaluations, our investigation revealed that his use of sources was by no means passive, but rather a complex process of creative reformulation. Thus, while the textual elements in al-Sharafī’s map production seem at first glance to be the work of a copyist, upon closer examination it becomes clear that his adaptation of sources is based on a complex methodology of realizing content-related goals combined with an explicit authorial intent. In addition to drawing on previous geographical knowledge, al-Sharafī also appropriated graphic elements from several Arabic sources as well as textual and graphic features found in geographical and cosmological genres and formats from other linguistic and cultural contexts. As we will see later, these primarily included references to Classical Antiquity (particularly Ptolemaic geography), to early modern European cartographical traditions, and to a lesser extent, to other cosmological traditions, such as the Indian one. Hence, his work constitutes an exemplary case of cultural translation.²² In this sense, we can describe al-Sharafī’s methodology as a multilayered quest for dynamic equivalence,²³ with the atlases being the target language in which semantic (and in al-Sharafī’s case, also semiotic) equivalents of a geographic treatise (as well as other literary, visual, and identity-defining sources) were sought and integrated.

10.3 Al-Sharafī’s Appropriation of al-Idrīsī’s Book

In preparing his atlases, al-Sharafī consulted other geographical, astronomical, and religious texts and relied on oral forms of knowledge;²⁴ however, al-Idrīsīʼs above-mentioned book was the source he relied on the most.²⁵ The manuscripts²⁶ comprising this book are a geographical compendium containing a series of small sectional charts and a description of a rectangular world map which, if ever executed, did not survive.²⁷ Later copies of al-Idrīsī’s work include a circular world map that is not described in the text.²⁸ Much later, in the twentieth century, Konrad Miller (d. 1933) reconstructed a full-scale rectangular world map based on six surviving manuscripts of al-Idrīsī’s sectorial charts.²⁹ Recently, the Factum Art Foundation also reconstructed the sectorial charts to form a whole rectangular chart, basing its work on the complete copy preserved in the Oxford Bodleian Library.³⁰ In this respect, the importance of ʿAlī al-Sharafī and his son Muḥammad resides precisely in the fact that they are the first known mapmakers to have reconstructed al-Idrīsī’s sectorial charts and adapted them to their own cartographic products, respectively, their 1579 and 1601 world maps.³¹ While beyond the scope of our paper, the question of how al-Sharafī gained access to al-Idrīsī’s work is a pertinent one.³² None of the copies of the Nuzha catalogued so far was copied in North Africa. However, there is evidence of its use in Ifriqiyya (modern-day Tunisia) since the thirteenth century by authors like Ibn Saʿīd al-Maghribī (d. 1286) and Ibn ʿAbd al-Muʾmin al-Ḥimyarī (d. after 1325).³³ In addition to its value for the study of geography, there is evidence that al-Idrīsī’s work was used in some Morisco circles as a tool for Arabic language instruction and geographical consultation, although it is hard to determine how widespread this use was.³⁴ For our purposes, however, what matters is that al-Sharafī made use of the text of the Nuzha in multiple ways. In al-Sharafī’s work, the textual, iconographic, and cartographic elements are intertwined to the point of being almost interdependent.

10.3.1 Al-Sharafī’s Authorial Strategies

We will now proceed to analyze al-Sharafī’s authorial (and traductological) strategies with regard to his appropriation of al-Idrīsī’s text. Our exploration will also take into account the ways in which the material characteristics of his opera (including matters of format) affected his decisions.

10.3.1.1 Format and Materiality

Materiality played a decisive role in al-Sharafī’s cultural translation strategies. In his use of sources, some material aspects shaped his decisions at the textual, iconographic, and cartographic levels. This is clear in his choice of the format³⁵ and in the material supports he used (cardboard, paper, and parchment), which help explain the substantial differences between the various renditions of al-Idrīsī’s quotations in al-Sharafī’s work.³⁶ The 1551 atlas, for example, is designed in accordance with the format and measurements of medieval Qur’ans from North Africa and al-Andalus. Format-related decisions also apply to the use of quotes from the Nuzha. These appear throughout his atlases, in particular in the explanatory texts accompanying the diagram of the celestial spheres (fol. 2^r) and the small circular world map (fol. 2^v–3^r) in the 1571 atlas, and at the top of the circular world map of the 1551 atlas (fol. 3^r).³⁷

The latter world map displays the most abridged version of what closely resembles one of the opening passages of the Nuzha, a recurrent passage throughout his work.³⁸ One possible explanation for this abridgment could be that al-Sharafī did not have direct access to any complete manuscript of al-Idrīsī’s work until later in his career, and may have been working with indirect quotations from other authors. Another perhaps more plausible explanation relates to al-Sharafī’s authorial strategies as shaped in part by the format of each individual work and the different concepts he employed for each of his atlases. This hypothesis is supported by other characteristics of the world map in question (see Fig. 10.1), which merge aspects of Idrīsīan-style mapmaking with that of others.

Fig. 10.1

Al-Sharafī’s atlas, 1551. BnF, Arabe 2278, fol. 3^r

The result of this is the creation, whether intentionally or not, of an original “Sharafīan” product, as opposed to a mere imitation of al-Idrīsī. Al-Idrīsī’s influence is seen, for example, on this map in the shape of the Black Sea and the Arabian Peninsula.³⁹ Both are ultimately derived from ancient Greek sources. Idrīsīan ancestry can also be discerned in the islands in the Indian and Pacific oceans.⁴⁰ Some of the islands in the Atlantic and the North Sea were already marked in the circular world map attached to some of al-Idrīsī’s manuscripts. Even the shape of the Atlas Mountains and the Nile sources might have been inspired by al-Idrīsī’s sectorial maps of North Africa. But it is more likely that the shape of the Atlas Mountains in al-Sharafī’s map was derived from a Majorcan portolan chart (see Fig. 10.2).

Fig. 10.2

Pedro Rosell, BnF, CPL GE C-15118 (RES), dated 2^nd half of 15^th century

The shape of the Caspian Sea, with its trifurcate contours, is already documented in a circular world map contained in a copy of Ibn Hawqal’s (d. after 978) (see Fig. 10.3) Kitāb Ṣūrat al-arḍ (The Book of the Picture of the Earth). Other features resemble more closely the work of other mapmakers, notably the design of Mount Qāf. This can be found in many earlier circular world maps, for instance those by Zakariyya al-Qazwīnī (d. 1283) and Ibn al-Wardī (d. 1457)⁴¹ (see Fig. 10.4).

Figs. 10.3

Ibn Hawqal’s world map, BnF, Ṣūra al-arḍ. BnF, Arabe 2214, fols. 52^v–53^r, dated 1445–56

Fig. 10.4

Ibn al-Wardī’s world map, Kharīdat al-´ajā´ib. BnF, Arabe 2193, copy dated 1597

In addition to features derived from different Islamic mapping styles, elements belonging to the Mediterranean portolan chart tradition, such as the lines for the thirty-two winds and the coastal lines of Syria and North Africa, are prominent in our author’s world map. (see Fig. 10.2), It is precisely these hybrid stylistic features that come together in a new product that distinguishes al-Sharafī’s politics of translation and his individuality.

10.3.1.2 A Sharafīan or an Idrīsīan Text?

The richest and most complex aspects of al-Sharafī’s appropriation of the Nuzha can be discovered at the textual level. A comparison of the different surviving manuscripts of this work reveals that, while there are indeed some minor differences between them, none of them corresponds verbatim to al-Sharafī’s rendition of it. While al-Sharafī draws from it, he makes—in addition to the usual copying mistakes—significant abridgements and adaptations of certain passages. Mistakes occur, for instance, when words with the same ending appear in subsequent or nearby lines. This led the scribe (al-Sharafī or his collaborators) to accidentally skip a few words or even entire lines of text. In at least one case, discussed below, the multiplicity of Arabic technical terminology, which came about because scientific texts were translated from different cultures between the eighth and the early eleventh century, seems to be the reason why any given term is sometimes replaced by a different one. In other cases, exemplified below, the alterations are clearly deliberate, even if they only consist of a single word.

These modifications do not constitute consistent patterns within the text, but rather reflect general tendencies of al-Sharafī’s working practice.⁴² In the adaptation of the above-mentioned excerpt describing the characteristics of the Earth, there is a reference to an allegory of possibly Indian origin, in which the Earth is described as a cosmic egg floating in a receptacle of water.⁴³ Al-Idrīsī’s text (like all previous texts containing variations of this passage) refers to the “belly of the celestial sphere” (jawf al-falak) and to the “belly of the egg” (jawf al-bayḍa). In the second case, al-Sharafī replaced the word “belly” (jawf) with the word “middle/midst” (wasaṭ), writing “jawf al-falak wa-wasaṭ al-bayḍa”.⁴⁴ By al-Sharafī’s time, the standard term in the astral sciences was no longer “belly” but “middle”. The lack of consistency in his usage of the two terms suggests that wasaṭ was indeed such a common technical term that he replaced the older term simply as a matter of course. This hypothesis is supported by the fact that other sources referred to by al-Sharafī also use the term wasaṭ—from the astronomical work Risāla by al-Jaghmīnī (fl. 13^th century) to the work of the Andalusi Sufi Ibn al-ʿAbbād (d. 1389/90),⁴⁵ and the geographical work of Ibn al-ʿAṭṭār (d. c. 1470).⁴⁶

In other passages, technical information included in al-Idrīsī’s work is omitted. For instance, when the above-mentioned passage of the Nuzha⁴⁷ refers to the measurements of the equator and other terrestrial and astronomical information, it presents a detailed list of measurement units, including the Persian farsakh, the iṣbaʾ (‘digits’), and the ḥabbat al-sha’īr (‘barleycorn’).⁴⁸ Al-Sharafī dispensed with all of them in his rendition of the quotation in his atlas of 1571.⁴⁹ He also omitted the description of the equator as “the longest line on Earth just as the ecliptic (minṭaqat al-burūj) is the longest line in the heavens”.⁵⁰ The quest for textual sobriety might be the underlying reason for such abridgements. This characteristic is neither capricious nor merely a question of authorial taste, but was dictated by the nature of his cartographic practices. For example, when al-Idrīsī’s wrote the above-mentioned passage describing the characteristics of the Earth, he was introducing the notion that the Earth could be divided into seven climes.⁵¹ Here, al-Idrīsī explained that the climes were divided not by natural lines but by man-made ones, and that in some of them mountains were more prevalent, whereas cities and fortresses could be found in others.⁵² Al-Sharafī’s decision to delete this entire passage from the text accompanying the circular world map of the atlases (which in the case of the 1551 atlas appears above the map)⁵³ can be explained by the fact that neither atlas contains climatic divisions.⁵⁴ As we will see in more detail later, this can be considered an example of carticity in al-Sharafī’s practice, in the sense that the text reflects the characteristics of the map it accompanies. As such, it is also an example of medium-based translation, since the abridgement of the text is designed to fit the visual elements (and the format) of the overall multi-media product of which it is part.

10.3.2 Creating a Mediterranean and Maritime Product

Al-Sharafī’s efforts to simplify, streamline, and reduce to essentials the textual information appropriated from al-Idrīsī’s work went hand in hand with another goal that he sought to achieve. The structure of the atlases highlights the seas, especially the Mediterranean and the Black Sea, as the focus of his work, a structure predetermined by his sources, namely charts labelled in Latin script, which he transliterated into Arabic. To avoid a major conflict between this focus and his use of excerpts from al-Idrīsī’s book, al-Sharafī had to translate the passages borrowed from a universal geographical context into texts applicable to the only two seas of interest to him and his potential customers, who are likely to have been elite families from Qayrawān and Sfax. Thus, on the one hand we have the translation (through transliteration) of a European portolan chart into his own chart; and on the other, the translation (through selection and abridgement) of a universal geography into a series of regional sectional charts.

Perhaps surprisingly, al-Sharafī did not use the Nuzha in his explanatory notes accompanying the sectional sea charts of the Mediterranean and the Black Sea. In these cases, he drew from “al-Hamadhānī”, a likely reference to the Persian historian Rashīd al-Dīn (d. 1318).⁵⁵ This alone indicates that his textual choices were not random, but conscious and systematic.

In contrast, let us now consider the sections in which our author did use the Nuzha. Drawing on the concept of “carticity”, we have seen in Sect. 3.1.2 that al-Sharafī chose the Nuzha as the best work to construct an explanatory preface to the seas shown on his small circular world maps, thus stressing the interdependence between text and image. He adapted (translated) the essential information on the seven seas⁵⁶ contained in the Nuzha—the origin and beginning of each sea, its length and width, and the number of its islands –⁵⁷ in such a way that the attention of the reader (and viewer) is drawn towards the Mediterranean, about which he gives the most information. The sequence of the seas in the text corresponds to the order in which they are shown on the circular world map. It begins with the China Sea, passes through the Persian Gulf and the Arabian Peninsula, and then continues through to the Atlantic Ocean and the Strait of Gibraltar, finishing in the Strait of Istanbul and the Caspian Sea.⁵⁸

It is in this process of adapting the text of al-Idrīsī’s universal geography into a Mediterranean atlas that al-Sharafī’s authorial persona finds its most elaborate expression. While throughout his work al-Sharafī retains quotations of passages from the Nuzha that make reference to different parts of the world, a close examination of his textual selection throughout his opera makes it clear that he favours themes, motifs, and cultural and geographic references linked to the Mediterranean. By this we do not only mean references to cities in the Mediterranean region, but also more specifically to imagery related to the sea itself. This allows al-Sharafī to ground his works in a localized environment. We could even call it a strategy of “domestication” of al-Idrīsī’s work. In translation theory, this term has been understood by Lawrence Venuti as a process through which the text in the target language is constructed so that it sounds familiar to the target audience, eliminating from it as far as possible the foreignness of the text in the source language.⁵⁹ By keeping the Mediterranean elements of al-Idrīsī’s work, while removing references to cities, seas, and other geographical landmarks from other regions, al-Sharafī domesticates al-Idrīsī’s work in order to create a new product rooted in a local context.

Furthermore, this process of domestication was fostered by yet another set of elements from the portolan atlases and charts that shaped al-Sharafī’s textual decisions. Given his emphasis on the seas, sometimes even at the expense of coastal—let alone inland—cities, he naturally opts for terms related to seafaring.

The 1571 atlas contains two very clear examples of the maritime nature of al-Sharafī’s creation. In it, our author quotes a passage from al-Idrīsī’s text in which reference is made to two main bodies of water, namely the Gulf of Venice (Khalīj al-Bundiqiyya) and the Black Sea (Khalīj Nītus).⁶⁰ When comparing al-Sharafī’s rendition to al-Idrīsī’s, we notice a major omission. Al-Idrīsī follows the reference to the Gulf of Venice with a geographical itinerary taking in the major coastal cities along it:

[…] it starts from the East from the lands of Calabria, from the lands of Rum and from the city of Otranto (Adhrant/Adhrantū), it passes by the northern side and turns westwards towards the land of Bari towards the coast of Sant Angelo (Shant Anjil) and then runs through the West and along the western shore towards the land of Ancona and passes towards the coasts of Venice and ends in the city of Aquileia (Īkalāya). And from there its shore in turn curves eastwards towards the lands of Croatia (Jarwāsiya), Dalmatia (Dalmāsiya), and Esclavonia (Isqalūniya) until it joins the Syrian Sea (al-Baḥr al-Shāmī).⁶¹

In a similar fashion, al-Idrīsī’s text follows the mention of Nītus with that of the cities of Trabzon and Constantinople.⁶² In al-Sharafī’s rendition, these detailed descriptions of the seas’ coasts and the listing of the cities along them are omitted, and he instead concentrates exclusively on the bodies of water in the atlases.⁶³

However, al-Sharafī’s reformulation of al-Idrīsī’s universal geography not only involves omitting references to the seas and gulfs. In other instances, our author substitutes his own, usually more condensed, rendition of the descriptions that follow the reference to a given sea. For example, after mentioning the eastern Mediterranean (Syrian Sea; al-Baḥr al-Shāmī), he deletes al-Idrīsī’s lengthy descriptions of the distances between its islands and his mention of the fact that it begins in what is considered the Fourth Clime.⁶⁴ In their place he inserts two brief passages adapted from the Nuzha, which are located in different places in the original work.⁶⁵ This demonstrates not only his focus on the Mediterranean but also the elaborate way in which he composed his texts. Of the sea in question he says:

The Syrian Sea begins at the Western-Surrounded Sea which is called the Dark Sea (al-Baḥr al-Muẓlim),⁶⁶ and only God knows what follows the Dark Sea. It is said that in this Western-Surrounded Sea there are 27,000 islands. The Syrian Sea begins at the Strait of Ceuta and [reaches] Syria, and for that reason it is called the Syrian Sea.⁶⁷

In this case, al-Sharafī replaces his source text with a semantically equivalent passage that fits both the needs of the format and the nature of the portolan atlas genre.

Through this strategy, al-Sharafī’s product can also be said to reflect a certain degree of “carticity”, as defined earlier as an intersemiotic translation process.⁶⁸ The structure of the text, with its sequences of seas and coastal towns, mirrors that of the small circular world map, thus creating a certain interdependency between text and image. Therefore, while the textual choices serve to domesticate the atlases, making them a specifically Mediterranean product, they also serve as a strategy of multi-media translation in which text and map are made to correspond.⁶⁹

Finally, through these same strategies of abridgement and adaptation al-Sharafī manages to square the circle, retaining al-Idrīsī’s authority as a solid source of high-level, widely accepted geographical knowledge, while creating his own persona as a successful translator and an inspired creator of a new maritime atlas that is recognizably rooted in his own cultural context.

10.4 Conclusions

Throughout this study we have seen how al-Sharafī’s decisions enabled him to construct a recognizable message that permeates his cartographic oeuvre. The model of the portolan atlas defined the way in which al-Sharafī adapted passages from al-Idrīsī’s text for his own new product, creating a textual support for his otherwise mostly visual (cartographic and iconographic) enterprise. His most common strategy for adapting the Nuzha so as to transform the atlases into a local product was the selective abridgement of quotations. But he also introduced changes into al-Idrīsī’s text that were generally neither random nor the result of copying mistakes. Moreover, the texts were modified to harmonize with the format and the visual characteristics of the atlases. On one level, the choice and modification of the texts were intended to respond to spatial challenges such as format and artistic style. On a second level, as we have seen, the texts were clearly selected, reformulated, and modified with semantic and epistemological concerns in mind.

Al-Sharafī’s rather frugal use of language was designed to communicate only what was deemed essential for each image or cartographic project. This is why he economized on references to measurements and technical and mathematical language and dispensed with references to the climes when the maps linked to the texts did not exhibit climatic divisions. These same authorial strategies helped create a relationship of interdependence between text and image, which can best be explained by the above-mentioned concept of textual “carticity”. In the examples of carticity alluded to by Italiano in his studies on Ariosto, the implication is that a new text is designed from scratch to mirror the image of a map. In our author’s case, however, what we see is the adaptation of a pre-existing text (the Nuzha), which itself was already intended to accompany a cartographic product, into a portolan atlas with its accompanying text. Thus, while carticity in and of itself already implies a certain kind of media translation, in that the text is designed to reflect the visual elements of the map, in al-Sharafī’s adaptation of the Nuzha the carticity of his text in relation to his own maps represents a higher layer of translation: that is, the translation/adaptation of al-Idrisī’s text into his own rendition is informed by the design of his own maps, which are themselves hybrid products, displaying features from different mapmaking traditions.

However, the most important aspect of al-Sharafī’s textual decisions relates, firstly, to his need to anchor his work within the broad intellectual tradition of geographical knowledge in Classical Arabic, and secondly, to his wish to create a localized Mediterranean product, by favouring references to sea over land, and to the Mediterranean region over other areas of the world. In stark contrast to some Ottoman mapmakers and geographers of his time, for instance, al-Sharafī shows no interest in including the Americas in his cartographic works, most likely because their inclusion would have been inconsistent with the Mediterranean focus of his work.⁷⁰ These two driving forces behind his textual and cartographic enterprise are the most clearly political aspects of his production, insofar as they complementarily allow al-Sharafī to present himself as a cosmopolitan figure able to quote from the masters of the broader Arabic tradition while at the same time fashioning himself as a North African and Mediterranean erudite and craftsman. It is precisely these politics of authorial self-fashioning that drive and enable al-Sharafī to position his work as distinct and authentic, while still allowing his sources to take center stage. In other words, al-Sharafī’s politics of translation aimed—in addition to embedding and transforming regional models into local cultural contexts—to cast his own persona as an author, mapmaker, book illustrator, and cultural broker, and to endow his final products with cultural and intellectual authority so that they would be recognized as familiar, valuable, legitimate, and attractive.