From the fifteenth through the seventeenth centuries, military engineers in the Mediterranean devised a new strategy for defending a city built on a peninsular site: a navigable canal was excavated through the neck of the landmass, severing the city from the coast and isolating it within the sea. In Building with Water: The Rise of the Island-City in the Early Modern Mediterranean, Elizabeth Kassler-Taub traces the development and dissemination of this overlooked urban type. She details how the “island-city” first emerged in the Adriatic and Ionian territories of the Venetian stato da mar and later swept across Spanish and Portuguese outposts in the western Mediterranean basin, where it was absorbed into a shared Iberian vernacular. By reconstructing the circulation of the island-city through this sprawling network of colonial frontiers, Kassler-Taub argues, we can chart an alternative path of architectural influence in the region, one that shifts our attention beyond the Italian and Spanish mainlands.
Military engineers dispatched to the Sicilian coastline in the first decades of Spanish Habsburg rule encountered a vulnerable landscape punctuated by aging fortifications incapable of withstanding the assault of modern artillery.1 In his Relazione delle cose di Sicilia of 1546, Viceroy Ferrante Gonzaga wrote that at the outset of his tenure in 1535, he found the island “very weak and open” to the near and present danger posed by Ottoman expansionist ambitions and the increasing corsair presence in and around North Africa.2 Gonzaga's successor, Don García de Toledo, echoed this sense of urgency in a letter of 1570, in which he described the pressing (“urgente, et urgentissima”) need to fortify ports and strengthen local infrastructure.3 Throughout the sixteenth century, Spanish efforts to remedy the state of Sicilian defensive systems concentrated on the construction of a nearly continuous chain of watchtowers and bastioned cities intended to gird the island against rising threats.4
The small city of Trapani, at Sicily's northwestern corner, was thought to be at particular risk given its proximity to North Africa. Located at the tip of a narrow tongue of land projecting from the coast, Trapani was also uniquely susceptible to incursions from both land and sea. In 1526, to counteract the city's topographical vulnerability, engineer Pietro Antonio Tomasello de Padua cut a sea canal along the existing stretch of walls fronting the mainland. A view of the city published in Rome in the mid-sixteenth century indicates the contours of this artificial waterway, which severed the peninsula—and with it, the urban core—from the terra ferma (Figure 1).5 In his Istoria di Trapani (1595), engineer-cum-chronicler Francesco Pugnatore observed that the canal “extends along the eastern walls of the city, such that the water of the sea, passing through, leaves the city and the castle—together—isolated.”6 The city was thus transformed into a fortified island, suspended just beyond the shore.
Tomasello's reconception of Trapani as an “island-city” was not an anomaly in the history of early modern military engineering. From the fifteenth through the seventeenth centuries, the practice of excavating defensive waterways across the isthmuses of peninsular cities gained currency across the Mediterranean. In this article, I trace the rise and dissemination of this urban type, arguing that the island-city was likely modeled on interventions in ancient settlements in the Greek archipelago and that it emerged as an early modern phenomenon in the Adriatic and Ionian territories of the Venetian stato da mar.7 Over the course of the sixteenth century, a wave of experimentation with the type swept Iberian outposts in the western Mediterranean basin, from North Africa to the Tuscan coast. By the turn of the seventeenth century, it had been fully assimilated into a shared Iberian vernacular of defensive design.
The circulation of the island-city type between these far-flung places might seem to rehearse a familiar narrative of architectural cross-pollination. Fortification technologies were rendered portable by itinerant engineers, who routinely left behind drawn and textual records as they traversed the Mediterranean. What sets this case apart, however, is the insight it offers into the role of colonial territories in developing defensive models within the larger region. By reconstructing the mechanisms that fueled the dissemination of the island-city, we can chart an alternative map of architectural influence in the Mediterranean—one that finds its center of gravity well beyond the familiar shores of the Italian and Iberian Peninsulas. Indeed, Sicily was an incubator for the development of this type throughout the sixteenth and seventeenth centuries, a fact that challenges the island's long-standing marginalization in the historiography of early modern architecture.8 In focusing attention on the exchange of architectural knowledge between understudied geographies, this article furthers contemporary disciplinary efforts to forgo the traditional dichotomy between “centers” of production and their “peripheries” in favor of a more flexible understanding of the cultural dynamics that shaped the period.9
Building with Water
The rise of the island-city demands that we rethink the status of landscape in early modern defensive design. Landscape had, of course, loomed large in the discourse of urban planning at least since Vitruvius's De architectura, which espoused a direct link between a city's site and its internal architectural logic. In the Vitruvian tradition, the orientation of streets was often predicated on the direction of winds in order to prevent unhealthy air from penetrating the urban fabric. In the case of fortified cities, this basic urban calculus intersected with the burgeoning science of warfare. The topographical challenges posed by coastal environments complicated the negotiation between city and site yet further: in a city with waterside ramparts, for instance, Vitruvius stipulated that the forum should be placed adjacent to the harbor; in an inland city, in contrast, the forum should be situated at the town center.10
Writing in the late fifteenth century, Francesco di Giorgio Martini expressed a distinctly Vitruvian attitude toward landscape. When designing the walled enclosure of a stronghold or fortress, he advised, the “prudent and skilled” architect must adapt the rules of building to suit “the nature of the site, adding and reducing and composing.”11 We find the same principle at play in Francesco de' Marchi's Della architettura militare, published posthumously in 1599. This treatise includes a series of illustrations of fortified cities abutting natural or artificial ports. In one of de' Marchi's imagined cities, a circular harbor interrupts the neat, radial organization of the streets like a bite from an apple; in another, a gridded city bends to meet the river bifurcating it, the parallel streets distorted to match the waterway's undulating contours (Figure 2).12 For both Martini and de' Marchi, topography determined urban morphology.
As Spain claimed coastal frontiers across the Mediterranean in the early decades of the sixteenth century, the elemental tension between landscape and design that had long defined military urbanism took on new urgency. Engineers were often forced to construct walls and bastions on unstable terrain, subject to the ebb and flow of shifting tides. These conditions prompted the development of diverse strategies for harnessing aqueous environments, from earthen embankments to large-scale land reclamation. Spanish builders readily drew on foreign expertise in hydrology. In the seventeenth century, engineers from the Low Countries were in demand across the Spanish empire. A Dutch engineer was tasked with the drainage efforts that carved the rigid urban plan of colonial Mexico City from the watery landscape of Tenochtitlan; similarly, a Flemish engineer fortified key Spanish outposts in the Mediterranean.13 As architectural historian Maurizio Vesco has shown, Spanish viceregal authorities on sixteenth-century Sicily instead demonstrated a preference for Venetian-trained engineers. Having cut their teeth in the urban canals and flooded marshes of Venice and the Veneto, such figures were well versed in what one writer, in 1533, called the method of “building in water.”14
That method was especially valuable on Sicily, where the extension of coastal cities' fortified perimeters beyond the water's edge required that wall and bastion foundations be laid in silt. Engineers in Palermo devised machinery to lower stone blocks onto the seabed and developed mortar capable of resisting the erosion caused by continual exposure to salt water.15 Tomasello, who transformed Trapani into an island-city, was likely conversant in such practices given his prior position as magistro ingigneri of Padua, a Venetian colony that experienced periodic flooding by the Brenta River.16 At midcentury, the engineer Bartolomeo Cascon, also of Venetian origin, was similarly charged with overseeing the construction of bastions on the muddy banks of a canal in Syracuse, on Sicily's southeastern coast.17
For Habsburg authorities scattered across the southern Mediterranean, Venice was synonymous with water. Following Charles V's successful campaign at Tunis in 1535, the Italians and Spaniards tasked with refortifying the nearby presidio, or military outpost, of Goletta found themselves mired in a messy building site that one engineer darkly described as nothing but “water and mud.”18 Perched on a narrow strip of land that divided the sea from a shallow lagoon fronting the city of Tunis, Goletta was battered by currents from both fore and rear. The Lombard engineer Ferramolino da Bergamo, then engaged in the refortification of Sicily, was commissioned to produce a series of designs for Goletta. His proposals proved unsuccessful, as they required walls and bastions to be constructed directly on the seabed. Following Ferramolino's departure from the presidio, engineer Francisco de Tovar begged Charles V to enlist the help of a Venetian, “because in Venice the foundations are made in water, as it must be done here … it thus would be advisable for you to send an official from there.”19
Despite its regional associations, the concept of “building in water” gained a foothold in Spanish architectural discourse by the end of the sixteenth century. Cristóbal de Rojas used similar language in his Teórica y práctica de fortificación (Madrid, 1598), where he described the proper “mode of building in the sea” and commented on the “inconveniences” of constructing a fortress or tower in aqueous conditions.20,Los veinte y un libros de los ingenios y máquinas, a hydrological treatise written by Juanelo Turriano that same year, similarly outlined how best to construct buildings “within the sea” and how to stabilize stone walls against the “great fury and impetuousness” of crashing waves.21 Teofilo Gallacini's early seventeenth-century treatise on the architectural geometry of the ideal port remains our primary touchstone for innovations supporting construction in coastal sites during the period. Like Turriano, Gallacini detailed complex systems of driven piles, wooden chains, and other apparatuses to support construction on seabeds and imagined fantastical devices to enable engineers to breathe underwater and “walk” across its surface.22 In the final years of the sixteenth century, Gallacini produced drawings for the new port of viceregal Naples, further demonstrating the importance of foreign hydrological technologies within a Spanish context.23
These construction strategies betray an essential pragmatism, one that defied rigid distinctions between a city and its preexisting site. If the Vitruvian tradition of urban planning placed architecture at the mercy of landscape, building in water quite literally breached its boundaries. Raising a stone wall against the tides may have been an innovation born of necessity, but it was also a powerful statement of architecture's potential to defy landscape. The island-city, exploiting the limitations of a peninsular site to defensive advantage, was nothing if not pragmatic. Yet the excavation of an artificial waterway represented a decidedly nonarchitectural solution to the problem of cities projecting into the sea. In this sense, the island-city should be understood as the result of a subtractive practice that shifted architecture's materiality from brick and mortar to landscape itself. To create an island-city was, in effect, to build with water.
Since antiquity, engineers have channeled through land and redrawn the topographical boundaries of coastal settlements, providing historical precedents for later, landscape-driven defensive design. Greek and Roman war chronicles are awash with references to the excavation of ditches—most often dry—along the stone walls of fortified towns or along the wooden ramparts and earthworks of temporary camps.24 Wet ditches, in turn, harbored vessels from storm or siege, or provided access between adjacent harbors.25 Artificial waterways were also excavated across coastal promontories. The earliest known example of this type of waterway is a defensive ditch in Vroulia, at the southern tip of Rhodes, datable to the mid-seventh century BCE.26 Also in the seventh century BCE, Corinthian colonists separated Leucadia (the island of modern Leukas, or Levkás) from the Acarnanian coast of Greece by digging a navigable canal across the isthmus joining the peninsula to the terra ferma.27 By the time of the Peloponnesian War, the canal had filled in with deposits of silt, forcing ships to be dragged across the isthmus. In the first century, Roman engineers recut the canal and constructed a stone bridge across it to connect Leucadia to the mainland.28 And in the fifth century BCE, the Achaemenid king Xerxes commissioned the Persian engineers Artachaees and Bubares to dig a similar canal across the mountainous peninsula of Athos. Herodotus characterized the canal as an exercise in vanity, rather than a necessity of war: “With no trouble they could have drawn their ships across the isthmus, yet [Xerxes] ordered them to dig a canal from sea to sea, wide enough to float two triremes rowed abreast.”29 Lysias, in turn, eulogized the act of “trenching Athos” as a statement of Xerxes's ascendancy over nature: “Despising alike the effects of nature, the dispositions of Heaven and the purposes of men … he forced a passage for ships across the land.”30
In each of these three cases—at Vroulia, Leucadia, and Athos—the canal's excavation was intended to provide expedient passage for ships, rather than to surround a peninsular city with water for defensive purposes. Nonetheless, these waterways were likely prototypes for the form of canal (fosso or fossato) that defined the early modern island-city.31
The practices of building in water and building with water—both modes of direct intervention in the natural topography—accord more closely with our contemporary discourse of “landscape urbanism” than with theorizations of early modern design. As James Corner and others have described it, landscape urbanism trades a simple opposition between natural topographical conditions and urban morphology for an awareness of their interpenetrations. This concept has informed recent discussions of the infrastructure of coastal cities, providing a platform for addressing the topographical mutability and architectural vulnerability of the urban waterfront.32 The significance of landscape-driven design in the Mediterranean from antiquity to the early modern period suggests that the current urge to reassess the architectural potential of the natural environment has deep historical roots. From an early modern perspective, this new language of landscape enables us to circumvent the traditional city–site binary and focus attention on the stakes (both practical and conceptual) of building at the littoral. It provides a model for incorporating the island-city—a type that, by its very name, speaks of landscape and urbanism in the same breath—into our architectural lexicon of the period.
Islands Real and Imagined
The arrival of the island-city on Sicilian shores in the first decades of the sixteenth century can be traced to earlier experiments in the eastern Mediterranean basin. An expansive theater of war, the region became a testing ground for new fortification technologies and, as such, represents a natural point of origin for the island-city. The emergence of the island-city in the eastern seas also coincided with a growing regional interest in spaces surrounded by water that crystallized in the genre of the isolario, or “book of islands.” First popularized in Venice, island books elevated landform to literary subject, producing a portrait of Italy's maritime hinterland that was at once cartographic survey and constructed historical narrative.33 The earliest volumes focused largely on the Aegean, the setting of naval battles such as the Ottoman sieges of Rhodes. Beginning with Cristoforo Buondelmonti's foundational Liber insularum archipelagi (1420), manuscripts and printed books created a taxonomy of Mediterranean and global archipelagoes.
Despite its pretense of empiricism, the genre was marked by a fluid and flexible understanding of topography. On the first page of his Isolario (1528), Benedetto Bordone declared his intention to describe not only islands—as the title of the volume would suggest—but rather “all of the islands and peninsulas of the world.”34 Tommaso Porcacchi's L'isole piu famose del mondo (1572) similarly gestured toward the relationship between the two topographical forms. A peninsula, Porcacchi wrote, is “almost an island, but is neither an island nor a continent”; an island is instead “separated and divided” from land and fully surrounded by water.35 Porcacchi prefaced his definitions with references to ancient authors. Strabo, he said, had argued that “all of the earth is an island, which looks out at the sea that surrounds it.” Pliny, too, maintained that all land—set within the expanse of the sea—“appears in the guise of an island.”36
This impulse to reimagine landmasses as islands was evidently felt across the eastern Mediterranean basin. In 1521—just five years before Tomasello transformed Trapani into an island-city—the Ottoman cartographer Piri Reis observed in his Kitab-ı Bahriye (Book of Navigation) that “Tripana [Trapani] is located on a low peninsula that almost looks like an island.”37 Piri Reis is known to have been familiar with the rhetorical conventions of island books. The maps that formed the focal point of his manuscript were based on those of the first printed isolario, published by Bartolommeo da li Sonetti in 1485.38
In the early modern imaginary, land not only had the capacity to masquerade as an island but also could quite literally assume its form. Throughout the period, Sicilian humanists debated claims by Diodorus Siculus and others that Sicily was once connected to the Italian mainland and had separated from its southernmost tip only after a violent earthquake or flood had forced the sea through the peninsula.39 In his Descriptio Siciliae, written in the mid-sixteenth century, Antonio Filoteo degli Omedei suggested that the division of these adjacent landmasses—“separated, sawed, cut, and divided from one another”—was literalized in the etymological origins of the island's name, taken by ancient authors to mean “to cut” (sicilire).40 Antonio's close contemporary Tommaso Fazello mused that during antiquity the earth was “disturbed and distorted.” Mountains rose up from valleys, islands surfaced spontaneously from the bowels of the sea, continents collided and fused, and waves carved new contours into the coastlines. It was only reasonable, Fazello remarked, to assume that Sicily was once attached to Italy.41
From the perspective of these beliefs in the mutability of topography, even a city could be mistaken for an island. Constantinople, Venice, and Tenochtitlan—each of which was famously surrounded by water—appear in the pages of island books, sandwiched between islands large and small. The urban core of Byzantine-era Constantinople occupied an anvil-shaped tract of land between the Golden Horn to the north and the Bosphorus to the south and east. As a cartographic view of the city in Buondelmonti's manuscript illustrates, a sinewy sea moat was excavated along nearly the full length of the fifth-century Theodosian walls, partially severing Constantinople from the terra ferma and making it almost an island-city (Figure 3).42 In contrast, a spare linear woodcut of Venice in Bordone's volume underscores that city's natural insularity. Framed by the terra ferma at left and a sequence of islands at right, the image sets the city adrift on the surface of the lagoon (Figure 4).43 For Porcacchi, too, cities that appeared “in the guise” of islands defied simple definition. He justified his inclusion of Constantinople among his description of “so many islands of the archipelago” by noting that the city is “all that an island is not” but is “nevertheless bathed on three sides by the sea.”44 Porcacchi's ambivalence aside, the first edition of his text is bookended by views of Venice and Tenochtitlan, with the “Isola, et Città di Vinetia” appearing first and the “gran Città e isola Temistitan” last.45 By identifying both cities as islands, Porcacchi collapsed the distinction between topography and urban space.
From Buondelmonti to Porcacchi, island books reflected fundamental changes in how topography was theorized and experienced. The historiographical tendency to view military architecture in primarily technical terms has left little room for recognizing how defensive innovations reflected broader shifts in cultural consciousness.46 Yet the concurrent rise of island books and island-cities suggests that this early modern awareness of landscape and its architectural potential resonated widely across the Mediterranean. In this sense, the isolario expressed the slippage between island and city in word and image at the same moment that the island-city was first realized in water and earth.
Between Southern Italy and the
Stato da Mar
Following the Ottoman siege of the port of Otranto in 1480, the Aragonese began an intensive project to strengthen the lapsed defenses of Puglia, the “heel” of the Italian boot. In the city of Taranto, located on a peninsula on the Adriatic coast, silt and detritus were cleared from a navigable sea canal first opened in Roman antiquity. Engineers widened the canal by cutting through the rocky land front of the peninsula, thus isolating the city from the terra ferma (Figures 5 and 6).47 In 1484, Venice seized the nearby city of Gallipoli and similarly cut a canal across the narrow isthmus connecting that city to the Puglian coastline.48 It has been suggested that the canal in Gallipoli was initially planned by Venetian engineers but only later executed by the Spanish, around 1500.49 However, contemporary fortification efforts on the Ionian island of Corfu indicate that this defensive strategy was endemic to the Venetian stato da mar. In the town of Corfu, engineers cut a navigable canal through the narrow neck of a rocky promontory where a citadel survived from the Byzantine period. The extant waterway divided the historic borgo from an expanding settlement on the terra ferma (Figure 7).50
These eastern Mediterranean island-cities likely served as early models for subsequent experiments on Sicily. In 1527—immediately after he severed Trapani from the western coast of the island—Tomasello called for the excavation of a canal across Syracuse's land front, which would have allowed a direct mode of communication between the flanking Porto Grande and Porto Piccolo (or Marmóreo).51 His proposal to make Ortygia—the main body of the Syracusan peninsula and the site of its historic settlement—into an island-city was not executed until the 1670s. Tomasello was, however, successful in isolating the Castello Maniace, a fortified complex at the tip of the peninsula, from the main body of the landmass.52 We might liken the Maniace to the Venetian type of “island-fort” (scoglio-fortezza), which converted a natural outcropping into a defensive machine.53
There was powerful precedent for the translation of defensive technologies from Puglia to Sicily. The engineer Antonello da Trani, employed on Sicily from 1518 to 1523 by Viceroy Ettore Pigantelli, was previously involved in the fortification of cities across Puglia, including Otranto, Barletta, and his native Trani. This southern Italian axis of communication was supported in part by familial patronage networks. In 1521—in the midst of his tenure on Sicily—Antonello was commissioned by Ettore's son Fabrizio, then prior of the knights, to design a fortress at Molfetta, near Bari.54 Puglian fortification technologies also formed part of the architectural inheritance that Antonello left to Tomasello, his successor on the island. The monumental round towers that Tomasello frequently built on Sicily were likely first introduced to Puglia in the late fifteenth century by Francesco di Giorgio, under the aegis of the Aragonese crown.55
These exchanges belie the dominant perception of Sicily's geographical allegiances during the period. The consolidation of Spanish control is thought to have forged an architectural link between Sicily and the Iberian world, pivoting the island westward, away from its Greco-Byzantine past. Yet Sicily remained in close contact with sites across the eastern Mediterranean throughout the early modern period and, as a result, was exposed to architectural innovations crisscrossing the region. In fact, all three engineers who oversaw refortification of the island in the first half of the sixteenth century—Antonello, Tomasello, and Ferramolino da Bergamo (whom we previously encountered in Goletta)—spent their formative years in Venetian milieus. Prior to his arrival on Sicily, Antonello served as capitano generale dell'artiglieria of the Venetian Republic from 1510 to 1512.56 Both Antonello and Ferramolino were trained by Gabriele dei Tadini da Martinegro, who played an instrumental role in the fortification of Venetian territories before he entered the Spanish service.57 Throughout their careers, Antonello, Tomasello, and Ferramolino can be placed in Venetian outposts as far east as Crete. These engineers thus entangled Sicily in the ties that bound Venice to its own rapidly expanding maritime empire.
Ferramolino's oeuvre offers further evidence of the circulation of architectural knowledge between Sicily and the Venetian stato da mar. In charge of fortification efforts on Sicily from 1537 to 1550, Ferramolino first arrived on the island from the Peloponnese port of Corone (Koroni), which had returned to Venetian control following a brief period in Ottoman hands.58 In 1538, at the outset of his tenure on Sicily, Ferramolino traveled to Dalmatia, along the eastern coast of the Adriatic, where he contributed to the defenses of Ragusan Dubrovnik.59 A letter the engineer sent to Ferrante Gonzaga, then viceroy of Sicily, indicates that he was also present on Corfu that year, where he would have been exposed to ongoing experimentation with the island-city type.
Between October 1537 and February 1538, the Veronese engineer Michele Sanmicheli widened the contrafossa that separated Corfu's fortified core from the terra ferma and constructed a scarped seawall along its banks.60 Shortly thereafter, in 1545 and 1555, the waterway was outfitted with a pair of monumental bastions executed by the military engineers Giulio Savorgnan and Marc Antonio Martinego (Figure 8).61 We can trace this basic formula for the fortification of an isthmus—a waterway flanked by a pair of bastions—to Sanmicheli's previous experience in the Dalmatian city of Zara (Zadar), also under Venetian control. There, he excavated a navigable canal (foša) across the isthmus of the peninsula occupied by the city (Figure 9).62 Begun in 1534, the canal was completed three years later, along with two angled bastions that controlled its points of entry. A heavily classicizing gate designed by Sanmicheli's nephew, Gian Girolamo, further dramatized the city's isolation from the terra ferma.63 These Adriatic and Ionian island-cities suggest that, in its infancy, the type was very much a Venetian phenomenon. It was Sicily's investment in Venetian architectural knowledge that laid the groundwork for the type's subsequent appropriation by engineers in the service of the Spanish crown.
Sicily and Malta shared a precarious position in the volatile waters between Italy and North Africa. After the fall of Rhodes in 1522, Charles V relinquished Malta and the adjacent island of Gozo to the displaced Hospitaller Knights. The diplomatic arrangement, fraught by haggling between the knights' grand master and the imperial court, was intended to gird Spanish territories against Ottoman expansion.64 The close contact between Sicily and Malta, coupled with the islands' defensive interdependence, encouraged the exchange of engineers and ideas.65 Period correspondence indicates that construction on Malta was overseen by the Sicilian viceroy, who, in conjunction with the grand master, wielded final approval on designs.66 In the midst of Palermo's refortification, Tomasello was sent to the Maltese city of Mdina by then viceroy Pignatelli to survey the architectural situation and produce a drawing of the royal castle and the city itself. The drawing was presented to the viceroy upon Tomasello's return to Sicily, indicating the depth of Pignatelli's investment in construction efforts on the island.67 Ferramolino similarly made two trips to Malta, in 1535 and 1541. There, he strengthened the defenses of the fortress of Saint Angelo, situated at the tip of the small peninsula of Birgu, by excavating a canal below sea level to shelter galleys from enemy fire.
The island-city type likewise circulated between Sicily and Malta, where the fortification of peninsular sites was a common concern. Early plans for the Maltese city of Valletta, constructed ex novo on the Sciberras Peninsula, included the excavation of a dry ditch from harbor to harbor across the isthmus—a final line of defense against a potential Ottoman siege (Figure 10).68 A letter sent from the grand master of the knights' order, Jean de la Cassière, to Phillip II in August 1576 reported that the engineer Scipione Campi had arrived in Valletta earlier that summer. Campi, who impressed the grand master with the “valor and experience” he showed during his visit, recommended widening the ditch and strengthening the site's existing bastions.69 The engineer was also in Syracuse that year, and his proposals for the two cities were likely shaped by one another. In Syracuse, he focused on the fortification of the land front, where in 1566 a small fortress had been constructed on the former site of the Castello Marchetti. The complex featured two city-side bastions in addition to a land-side pair constructed during Gonzaga's viceregal tenure.70
With the completion of Campi's plan in 1577, the city-side bastions and the adjacent portion of the fortress were demolished, and two new bastions were constructed beyond the old enceinte. As such, the late sixteenth-century defensive system of the Syracusan peninsula consisted of two parallel lines of fortification intended to protect the peninsula against assaults from either land or sea. The most significant dimension of Campi's 1576 proposal—the excavation of a waterway below sea level to divide the city from the terra ferma—remained unexecuted (Figure 11).71 In the same year, the engineer suggested cutting a nearly identical waterway through the peninsular city of Augusta, just north, although work on Augusta's fortifications was delayed in favor of concentrating resources on Syracuse.72 Tiburzio Spannocchi dismissed the depth of Campi's planned canal in Syracuse as unviable, citing the engineer's lack of experience in “executing works in water.”73 Spannocchi's own design for the fortification of Augusta nonetheless suggests that he recognized the value of this form of waterway. In his Descripción de las marinas de todo el reino de Sicilia (1578), an atlas of the Sicilian coastline commissioned by Philip II, Spannocchi recommended the excavation of a canal in Augusta, thus echoing Campi's designs for both cities. An accompanying illustration depicts Augusta severed from the Sicilian coastline, although the waterway was not executed because of the rocky terrain (Figure 12). An anonymous proposal for Augusta produced in the late sixteenth century once again called for cutting across the isthmus, attesting to the continued popularity of Campi's vision for the city.74
Efforts to transform Syracuse and Augusta into island-cities only intensified in the seventeenth century as connections between Sicily and Malta were forged anew. Interest in the defensive potential of peninsular sites on both islands—spanning nearly two centuries—made the region fertile ground for the development of the type. The island-city's localization in the southern Mediterranean should be understood, however, against the backdrop of a larger constellation of Iberian outposts where fortification efforts similarly revolved around cities projecting into the open sea.
Just as engineers trained in Venice and the Veneto facilitated the early dissemination of the island-city between the stato da mar and southern Italy, itinerant Italians who moved among the architectural circles of João III, Charles V, and Philip II popularized the type in both Spanish and Portuguese contexts.75 The transformation into an island-city of the Portuguese stronghold of Ceuta, located across the Strait of Gibraltar from Spain, was especially influential for the later adoption of the type in Iberian territories. Ceuta, which occupied the narrow neck of Morocco's Almina Peninsula, was bounded by terra ferma to the west and the looming profile of Monte Hacho to the east.76 The site's natural defenses were initially reinforced by a system of ditches. In the eleventh century, the Andalusian geographer Al-Bakri mentioned the presence of a dry ditch across the isthmus that isolated the city from the terra ferma. By the time of the Portuguese conquest in the early fifteenth century, that ditch was one of four distributed across Ceuta's eastern and western flanks.77 Although it was dry, it anticipated the reconception of the peninsula as an island.78
In 1541, the engineer Benedetto da Ravenna arrived in Ceuta in the company of his Portuguese counterpart Miguel de Arruda.79 Benedetto's service to the Portuguese crown followed almost three decades of work in Habsburg territories. Following his involvement in the campaign at Tripoli, he was appointed royal engineer of the Kingdom of Naples in 1511.80 After his eventual departure for Spain, Benedetto traveled widely, participating in the defense of Rhodes in 1522 and the invasion of Provence in 1524, alongside Tadini, who trained two of the architects responsible for the refortification of Sicily. It was Benedetto's knowledge of the defenses of Spain's North African outposts, gleaned during a journey to survey the region in 1534, as well as his experience in the modern technology of bastionization, that likely prompted João III to enlist his help inspecting Portuguese fortifications along the Moroccan coast.81 Benedetto's proposal for the refortification of Ceuta, outlined in a report sent to the imperial court in Lisbon, entailed the construction of a bastioned system along the perimeter of the city, including a pair of monumental angled bastions facing the terra ferma.82 Yet the defining Portuguese intervention in the city's defenses during Benedetto's tenure was topographical rather than architectural. Between 1541 and 1549, the existing dry ditch was converted into a broad, navigable canal that severed the city from the mainland. The still-extant waterway, which provided secure passage for ships between the Atlantic and the Mediterranean, cuts a dramatic profile along the Muralla Real, the modified line of fortifications at the city's land front (Figure 13).83
During the first half of the sixteenth century, the practice of severing peninsular cities or fortresses from the terra ferma spread across the Portuguese colonial world, from the Mediterranean to South Asia. Period sources indicate that Ceuta was a direct inspiration for Diu, located at the tip of a small peninsula on India's northwestern coast. Diu was occupied by the Portuguese since 1535, but it was not until 1546—when the excavation of the navigable canal in Ceuta was well under way—that a waterway was cut through sandstone at the base of the peninsula according to plans by the engineer Francisco Pires, thereby rendering Diu an island-city (Figure 14).84 The Portuguese had experimented with the fortification of peninsular sites in India since the beginning of the sixteenth century. Of particular note is Cananore (Kannur) on India's southwestern coast, which was officially absorbed by Portugal in 1505. As a defensive measure, a waterway was cut through the base of a small peninsula located a short distance from the town (Figure 15). The undertaking, which isolated a Portuguese fortaleza-feitor (fortress-warehouse) within the sea, was copied shortly thereafter in the Sri Lankan port of Colombo.85
Given the fluid movement of engineers between Iberian territories, knowledge of these Portuguese island-cities likely shaped subsequent interventions in Spanish outposts. The fortification of the presidio of Mazalquivir (Mers-el-Kébir), occupying a rocky promontory on the coast of modern Algeria, was marked by these trans-Iberian exchanges. Leonardo Turriano, who penned the definitive account of the fortification of Mazalquivir and the nearby presidio of Oran in November 1598, rose to prominence in the architectural circles of Philip II after making his name in the imperial court in Lisbon and contributing to the fortification of the Canary Islands, then under Portuguese control.86 When Turriano arrived in Mazalquivir in 1594 to survey the presidio's fortifications, he did so armed with broad knowledge of the Portuguese colonial landscape. Here, too, the legacy of early experiments in the eastern Mediterranean remained close at hand. Like the engineers responsible for Sicily's refortification, Turriano was familiar with the Venetian fortifications of Dalmatia. In his atlas of the Canaries, he referenced San Nicolò in Šibenik, a scoglio-fortezza attributable to Gian Girolamo Sanmicheli.87
Engineers never succeeded in severing Mazalquivir from the terra ferma, yet period accounts of the presidio's fortification reveal it as a flash point for debates around landscape-driven approaches to defensive design. The first phase of construction in Mazalquivir can be dated to between 1563 and 1569, when engineer Giovanni Battista Antonelli razed and reconstructed an existing fortified complex on the site.88 Antonelli subsequently returned to Mazalquivir from Spain in 1574, in the company of Vespasiano Gonzaga, a general and military engineer.89 Turriano, who detailed the disputed plans for the defense of Mazalquivir in his 1598 manuscript, lambasted Antonelli's interventions. Along with Vespasiano, he argued that the engineer had made the presidio even more vulnerable to the assault of enemy artillery by leaving the peninsula's tip devoid of any fortification.90 Antonelli had instead excavated a ditch across the rocky land front and fortified it with two lateral bastions. Vespasiano felt that Antonelli had blindly implemented a bastioned system without regard for the peninsula's particular topographical conditions.91 He proposed excavating the existing ditch below sea level and expanding the fortified enceinte such that it would hug the peninsula's rocky contours (Figure 16).92 By his own estimation, this would trade a moat for the sea itself. “The deep sea,” he argued, “is the greatest moat in the world.”93 Turriano characterized Vespasiano's designs for Mazalquivir as examples of “natural fortification.”94 We might liken the engineer's approach to the practice of building with water: at Mazalquivir, Vespasiano wrote, “We should trust more in the sea” than in brick-and-mortar defenses.95
Vespasiano's and Turriano's criticism of his work at Mazalquivir aside, Antonelli evidently shared his contemporaries' belief in the sea's defensive potential. Antonelli's Epitomi delle fortificationi moderne, written in Toledo between April 1560 and March 1561—just two years before his arrival in Mazalquivir—contains what may be the first direct reference to the island-city in period literature: “The maritime sites that are strong are those that are surrounded by harborless, impetuous, and tempestuous seas or by one part or the other disjointed from the land by canals.”96 The manipulation of waterways (both natural and artificial) was Antonelli's claim to fame. In 1529, he proposed the excavation of a canal in Panama; later, in the 1580s, he was at the forefront of a large-scale imperial initiative to render Spain's rivers navigable.97
Subsequent interest in the island-city type again passed back and forth between texts and building sites. At Cádiz, a peninsular city along Spain's southwestern coast, a wide, navigable canal was cut across the isthmus, transforming it into a true island-city (Figure 17).98 A woodcut included in Rojas's treatise of 1598 suggests that the engineer may have had a hand in the excavation of the canal. The image depicts an unnamed fortified city separated from the terra ferma by an artificial waterway (Figure 18). Scholars have identified the illustration as a reimagining of Cádiz, then bounded at its land front by a dry ditch.99 Rojas, who first arrived in Cádiz in 1589, was closely involved in its fortification. In 1597, he consulted on the defensive system of Ceuta, which had passed in 1581 from Portuguese to Spanish control.100 There, Rojas would have encountered the Portuguese-era canal, which represents the most likely source for his reimagining of Cádiz as an island-city. The publication of his treatise in 1598—coinciding with the date of his return to Cádiz—immediately followed his visit to Ceuta.101
Rojas was not the only engineer in Cádiz interested in the island-city type. Several others who had contributed to the fortification of Mazalquivir were involved in construction efforts there, including Vespasiano Gonzaga, whose initial design for fortifying the Algerian presidio's land front included a ditch excavated below sea level.102 Rojas inherited responsibility for the fortification of Mazalquivir from Spannocchi, who similarly proposed cutting a canal through the isthmus of Augusta in his atlas of 1578.103
Over the course of the sixteenth century, the island-city was made mobile by these itinerant engineers, whose overlapping paths between Spanish and Portuguese territories in the western Mediterranean turned their defensive experiments into a shared vernacular. In the years after Cádiz was severed from the terra ferma, that vernacular resurfaced on Sicily, where the type was first adopted in a Spanish milieu.
Syracuse and Augusta reemerged in the late history of the island-city. Mirroring designs of the 1570s by Campi and Spannocchi, the Tuscan engineer Giovanni de' Medici in 1640 proposed the excavation of a wide waterway across Augusta's land front, with a ravelin positioned at its midpoint (Figure 19). As was the case in the sixteenth century, the project was not completed; a plan of the city made by the Sicilian engineer and cartographer Francesco Negro in collaboration with Carlo Ventimiglia for their Atlante di città e fortezze del regno di Sicilia (1640) shows the isthmus still intact (Figure 20).104 That same year, de' Medici was also in Valletta, where a dry ditch and additional outworks of his design were constructed across the land front; he was also in Syracuse, where he likewise suggested the excavation of a waterway along that city's land-side bastions.105
De' Medici's approach anticipated the transformation of Augusta and Syracuse into island-cities some decades later.106 In 1671, the Flemish engineer Carlos de Grunenbergh began working to modernize the defenses of Sicily.107 His plans for Syracuse and Augusta, produced shortly after his arrival on the island, were sent to the imperial court in Madrid for approval. In Syracuse—once again prioritized over Augusta—work progressed swiftly and was largely completed by 1673.108 Grunenbergh's design for Syracuse, consistent with that of his predecessors, comprised parallel lines of fortification bounded by waterways that together flanked the isthmus.109 By excavating a broad darsena along the existing city-side bastions, Grunenbergh fully suspended the city within the sea (Figure 21).110 Contemporary observers mythologized the city's transformation: one late seventeenth-century account declared Syracuse a “peninsula made into an island by the art of war.”111 In his Memorie patrie per lo ristoro della città di Siracusa (1791), the local humanist Tommaso Gargallo similarly lauded the “famoso taglio” (famous cut) that made the city an island.112 That deceptively simple cut suggested a powerful return to origins. Ancient geographers believed the city was once a natural island off the Sicilian coast; according to Ovid, “There was a time when Ortygia floated on the waves, now it is fixed.”113
In 1682, a nearly identical canal was excavated across the Augustan isthmus (Figure 22).114 The completion of the waterway at Augusta immediately followed Grunenbergh's involvement in the fortification of Valletta. In January 1681, he briefly left Sicily for Malta, where he recommended modifications to the land front of the Floriana fortress. A letter sent from Messina in 1682 concerning Valletta's defenses indicates his awareness of ongoing construction efforts there even after his return to Sicily.115 Sensitivity to topographical conditions was a common thread of Grunenbergh's designs for both Valletta and Augusta. Like sixteenth-century proponents of “natural fortification,” he argued that Valletta's defenses had been constructed without regard for the terrain.116 He proposed adding a lower enceinte fronting the Grand and Marsamxett Harbors and stipulated that the walls be constructed in stone capable of withstanding continual spray from seawater. In Augusta, too, the bastioned enceinte was in direct contact with the sea; as a result, Grunenbergh was forced to stabilize the foundations with wooden piles driven into the seabed.117 His fluency in these construction technologies can be ascribed to his origins in the Low Countries, where—as was the case in Venice—engineers were steeped in the practice of “building in water.”
Carlos de Grunenbergh's prior exposure to Spain's territories elsewhere in the Mediterranean provides an alternate source for his knowledge of the island-city type. In 1669, he went along with his brother Fernando to the Tuscan Maremma, a coastal region formally annexed by Philip II in 1557 as the Stato dei Presìdi. Paralleling concurrent fortification efforts across Sicily, Malta, and North Africa, the Habsburg crown undertook an intensive campaign to modernize the region's defenses. Fernando completed several works in the city of Orbetello, including a new bastion, a powder magazine, and the land-side gate.118 The city, sited on a peninsula jutting into a lagoon, was flanked by two isthmuses, which together joined the presidio of Monte Argentario to the Tuscan coastline.119 In Orbetello, Fernando would have encountered a model of a true island-city. Before the Ottoman siege of 1646, an existing defensive waterway across the land front was expanded into a wide, navigable canal (Fosso Reale) (Figure 23).120 Given Carlos's involvement in the surrounding region of Grosseto, we can assume that he was also familiar with Orbetello's modified topography.
The seventeenth-century Fosso Reale was the culmination of earlier proposals to render Orbetello an island-city. Pietro Cataneo's plan of 1547 for modifications to Orbetello's defensive system entailed deepening the existing waterway across the isthmus. In his 1554 treatise, Cataneo acknowledged the particular strength of cities isolated by artificial waterways or the sea and claimed that surrounding Orbetello with water would make it second only to Venice, “the strongest city in the world.” Orbetello, he noted, was frequently called “little Venice” because of its topographical resemblance to that iconic “floating” city.121 Cataneo thus inscribed Orbetello into a literary tradition that cast cities as islands. In this sense, the seventeenth-century afterlife of the island-city speaks to the longue durée of the type, and to an enduring interest in urban spaces surrounded by the sea. The mutability of natural topography—its capacity to be reshaped and reimagined—had clearly entered the early modern bloodstream.
The island-cities discussed in this article by no means constitute an exhaustive list. Many of the engineers who surveyed and fortified Spanish territories in North Africa and Sicily were also sent to the crown's Atlantic and Caribbean colonies and likely brought knowledge of the island-city type with them to the Americas.122 Within the context of the Mediterranean, however, the rise of the island-city from the fifteenth through the seventeenth centuries subverts established paradigms of how—and, most important, where—architectural ideas developed. Indeed, the history of the island-city is a history of colonial frontiers, continually lashed by the tides of war. Only recently have we begun to view these frontiers as laboratories where designs were tested and ad hoc solutions devised. But rarely do we acknowledge the transmission of models between those places that linger at the margins of our gaze—whether from the stato da mar to Sicily or from one North African outpost to another. In its dissemination across the eastern and western seas, the island-city lays bare points of contact that transgressed territorial boundaries. As historians of early modern urbanism and military engineering increasingly look beyond the well-trodden landscapes of Italy and Spain, these neglected patterns of exchange may alter our understanding of the mobility of architectural knowledge in the Mediterranean.