Printer Friendly

The master-builder, the bureaucrat, and the practical soldier: protecting Alcacer Seguer/Qasr al-saghir (Morocco) in the Early Sixteenth Century.

On 18 October 1458 a Portuguese expeditionary force of around 25,000 men, led by king D. Afonso V, Infante D. Henrique, and other prominent noblemen, overwhelmed Qasr al-saghir (35[degrees] 50' N--5[degrees] 33'W), a small Moroccan coastal town located roughly midway between Ceuta and Tangier. The new conquest, all too diminutive given the resources marshalled to acquire it, nonetheless remained of great symbolic and diplomatic significance for the Portuguese Crown until the conquest of Tangier (Tanja) and Arzila (Azlla) in 1471. Even thereafter, Qasr al-saghir/Alcacer Seguer retained certain strategic importance as one of the three "anchors" of the northernmost stretch of the Portuguese military frontier in the Maghrib. The town nestled in the shadow of Cape al-Qasr, on the east bank of the estuary of modern-day Wadi al-Qasr (known to the Portuguese as Rio Canhete), which flows from the low hills of Jabal al-Gharbawiyin (elevation c. 250-390 m) northward toward the Strait of Gibraltar. The estuary bay was one of the few relatively sheltered anchorages between Tangier and Ceuta. By the 1530s, however, Alcacer stood high on the list of Portugal's expendable outposts in Morocco, and after a brief last-minute effort to render it defensible, the town was abandoned between mid-June and 16 July 1550. Its fortifications and main buildings were partially destroyed.

The locality was never permanently resettled after the Portuguese evacuation, and eventually became a prime "time-capsule" target for historical archaeology. To this day, Qasr al-saghir remains the only former Portuguese outpost where extensive excavations--greatly facilitated by the prolonged abandonment of the site--have managed to shed light on long-run processes of urban change and adaptation along the southern shore of the Strait of Gibraltar, the area that, having served until the fourteenth century as a Muslim gateway to the Iberian Peninsula, became by the 1470s the Algarve de alem-mar of Portuguese overseas expansion. From 1974 to 1981 Qasr al-saghir was the focus of a joint Moroccan-American archaeological project (Mission archeologique americaine au Maroc) under the direction of Prof. Charles Redman, supported by the State University of New York at Binghamton, the Smithsonian Institution, and other agencies. The excavated area covered some 18 per cent of the space enclosed by the town walls, 80 small soundings were made along the fortified perimeter, and further work was carried out in the citadel (the castelo built in several stages around the nucleus of the old Marinid double bent-axis Sea Gate, the Bab al-bahr). Our understanding of Qasr al-saghir's Muslim and Portuguese past would be much poorer were it not for the results of this sustained field work. (1)

The research mandate of the Mission archeologique necessarily reached beyond the narrow scope of the first half of the sixteenth century, the site's late Portuguese period. This, together with the interdisciplinary composition of the teams, the need to satisfy government officials and funding agencies at the end of each field season, and the need to establish basic frameworks in an archaeological context that had not been studied in a systematic fashion previously, called for difficult choices in terms of priorities, research strategy, and methodology, candidly discussed by Redman in his Qsar es-Seghir: An Archaeological View of Medieval Life. (2) None the less, especially where the fortifications of Portuguese Alcacer Seguer were concerned, the project remained marked from beginning to end by a disconnect between archaeology and archival investigation, a disconnect unwittingly spotlighted by Redman's confident portrayal of the extent to which excavations at the site benefitted from "information and insights gained from exhaustive studies of published historical records and investigations of archives in Rabat, Lisbon, and Madrid." (3) With regard to the Portuguese records, unfortunately, the research was nowhere near exhaustive. (4)


The issue goes deeper than a mere quibble about the scope of supporting research. Portuguese Alcacer Seguer is arguably one of the more meticulously documented sixteenth-century European fortified sites in North Africa. Its walls have been the object of successive royal instructions, building specifications, and detailed measurements. The documents capture most of the key stages marking the early sixteenth-century transformation of the town's perimeter, from a Muslim fortification built largely in the late thirteenth and fourteenth centuries and marginally improved by the Portuguese in the late fifteenth century, to an early modern colonial urban stronghold attempting to meet, albeit awkwardly and imperfectly, the new challenges and requirements of rapidly maturing gunpowder warfare. The excavation team would have certainly benefitted from a line-by-line dissection and field research corroboration of at least the two pivotal documents available in print before 1974, namely the 1502 "Instrucoes a respeito das obras da vila de Alcacer Seguer" (Arquivos Nacionais/Torre do Tombo [further AN/TT], Gaveta XV, 18-26, published in 1965) and the 1514 "record of survey" (auto de medicao) (AN/TT, Nucleo Antigo 769, first published by Vieira Guimaraes in 1916). (5)

Although the well-known sketch of Alcacer Seguer's couraca (shielded passageway) appended to the 1502 "Instrucoes" was chosen to illustrate the title page of Chapter Five in Redman's Qsar es-Seghir, the project's publications suggest that the relevant documents did not play a significant role in the excavation, dating, and analysis of Alcacer's fortifications. The following passage, referring to work within the citadel (castelejo or castelo), directed by James Boone, is emblematic of the problem:
   Excavations ... revealed that there were at least two distinct
   building phases and layouts.... Sometime during the occupation of
   the site the citadel was enlarged, creating the enclosure we refer
   to as the patio d'armes. It is difficult to determine with
   certainty the date for the building of these new fortifications.
   The logical date, however, would be sometimes during the first 50
   years of occupation [i.e. 1458-1508--note mine], when the
   Portuguese were subjected to intermittent attacks. (6)

An inspection of the published documents would have quickly established, however, that the "logical date" hypothesis was unsustainable, and that the most conspicuous additions were in fact the product of a crash three-year building program completed only in 1514. The impetus for the refortification of Alcacer Seguer were not "intermittent attacks" during the first fifty years of occupation, but a flashpan military crisis spawned by wrenching political changes in Morocco, an aggressive rise in Manueline military opportunism after 1505, and a rapid Moroccan adoption of up-to-date gunpowder weapon tactics. (7) On a structural level, the documents would have readily revealed, for instance, that the "enclosure" labeled for research purposes the "patio d'armes" was really the inner ward of a structure conceptualized by its builders as a bastion (the Baluarte da Praia), a fact that has direct bearing on any assessement of the tactical redesign of Alcacer Seguer's defences in the second decade of the sixteenth century. Finally, had the excavators discovered in the Torre do Tombo's Corpo Cronologico the 1509 "Regimento das obras de Alcacer" (unfortunately transcribed only in 1991, in an unpublished doctoral thesis), (8) they would have held all the basic tools to decipher more accurately the castelo's architecture, to frame a better targeted excavation strategy for the citadel area, and to associate uncovered structures with military and economic functions.

Matching the three documents with evidence on the ground would also have promoted a more refined chronological and tactical interpretation of the entire fortified perimeter, drawing attention to obscure but significant segments. Targeted soundings in such areas, time and resources permitting, might have helped to weed out from the project's overall ground plans of the Islamic and sixteenth-century enceintes tacitly embedded assumptions and omissions, which have been subsequently replicated in all derivative sketches and models of Alcacer Seguer. Since 1986, intriguing discrepancies between the plans and the written sources have tended to invite speculation about the function and dating of certain features, particularly on the town's west side, facing Wadi al-qasr, an exercise to which I contributed a few tempting but partially premature deductions in an earlier study. (9) In 2000-2, Pedro Dias' ground-breaking Arquitectura dos Portugueses em Matocos (1415-1769) reopened all the issues by taking a much closer look at the sources than any previous work, yet in the end failed to provide structurally satisfying answers. (10)

Having brought together all the main pieces of the puzzle--the excavation plans, the Torre do Tombo records, and photographs of the surviving fortifications--Dias stopped just short of drawing the necessary conclusions. His analysis juxtaposed the sources rather than cross-correlated them, with a tantalizing outcome. The Arquitectura's eighteen-page Alcacer chapter highlighted vividly the potential of the written sources and offered glimpses of structural features described or measured in the early sixteenth century but no longer conspicuous (i.e. not featured in the Redman plans and not readily visible on the ground), only to conclude that as far as the fortifications are concerned the 1514 auto de medicao simply corroborates modern visual inspections and the 1974-81 archaeological results--"confirma que o que hoje se pode ver e o que se deduz dos relatonos das excavacoes corresponde a esta empreitada [i.e. the building program culminating in 1514--note mine]") (11) This generalization, however, glosses over features of distinct relevance for understanding, on the one hand, the tactical thought of those responsible for the defence and architectural renovation of Alcacer Seguer, and on the other hand the practical, political, and bureaucratic constraints that weighed on the builders.

It seems only reasonable, therefore, to revisit the evidence in an attempt to produce a revised image of Alcacer Seguer's defences and of their evolution between the late fifteenth and early sixteenth centuries. The exercise is necessarily predicated on gauging the reliability of the extant sources--an issue that has never been given much consideration. The one comment that stands out in the literature, Dias' assessment of the design sketch for the couraca accompanying the 1502 "Instrucoes", is rather discouraging--"Devemos ter em atencao a ma qualidade do desenho, a incapacidade de quem o executou ..." (12) Axe the other records any better, particularly the 1514 auto de medicao covering most of Alcacer's fortified enceinte? Dias' Arquitectura casts no shadow of doubt over the auto, characterized as "de maior importancia". (13) Yet it also leaves unclear both the extent to which historians can indeed rely on this document and its companions, and the grounds for the author's opinion that the sources merely corroborate the archaeological observations.

Before tackling instrucoes and medicoes, stone and brick, towers and bastions, it might be useful, however, to outline the profile of the two key protagonists of Alcacer's early sixteenth century military reconfiguration--the relatively obscure soldier, entrepreneur and mestre de obras Francisco Danzilho, and the much better known mestre and Manueline era architect Diogo Boytac. (14) Both of them immigrants to Portugal, they exerted a significant influence over distinctive niches of Portuguese architecture. As far as Alcacer Seguer is concerned, Danzilho gave the town's Manueline fortifications their basic structure, still discernible despite the evacuation damage and centuries of decay. Boytac's close inspection of the works yielded the most detailed contemporary record of their dimensions and arrangement, in plan as well as in elevation. The uneasy working relationship between the two men, not very well contextualized thus far, has unfortunately rendered the historical evidence ostensibly more opaque by adding a legacy of possible situational bias to the usual challenges of dealing with the vagaries of sixteenth-century military and construction jargon and correlating specific architectural features in the documents and on the ground.

For the present purpose, the other figures associated in one way or another with the modernizing of Alcacer's defences do not require the same attention, and this is just as well, because the sources would not permit it in most instances. A short list will have to suffice. The array of master-builders and senior craftsmen included, in rough order of importance, Martim Lourenco, Sebastiao Luis, Diogo Barbudo, the mestre de pedraria Femao Gomes of Faro, and the carpinteiros Afonso Martins and Francisco Fernandes. (15) Among the officials, it is necessary to mention above all Pero (Pedro) Vaz, who for a number of years fulfilled alternatively the functions of vedor, recebedor, and contador of works at Alcacer, and supervised some of the early stages of the building program. (16) Next comes Martim (Martinho) Anes, the town's almoxarife in the early 1500s, (17) and then Fernando Vieira, the escrivao of works. (18) The dominant military personages looming in the background were the captains and governors of Alcacer D. Diogo de Sousa, (19) D. Pedro de Sousa, (20) and finally D. Rodrigo de Sousa. (21)

Francisco Danzilho ("de Anzilho", "de Lenzina", "Danzina" in the Portuguese sources, and "de Arcillo" in Spanish ones--possibly a misreading of "de Ancillo" or a scribal slip), was of Cantabrian origin, reputedly an hidalgo, from the small locality of Argonos, roughly midway between Noja and Santona, just west of Laredo and some 45 km from Santander. His patronymic seems to be rather a toponymic, likely derived from Dancillo (or Ancillo), the old fishing barrio of Argonos. The date of his move from Spain to Portugal remains unknown, but by 1510 he was already referred to in Castilian sources as a "vecino de Argonos residente en el reyno de Portugal." (22) Enjoying a solid regional reputation as a military engineer by 1502-8, he took part in the refitting of the fortifications along Portugal's eastern borderlands, in the wake of the initial grant of funds by the Cortes of Lisbon in 1502.

Danzilho's work is attested above all in Miranda do Douro (pentagonal bastion attached to a new section of the fausse-braye), 23 in Almeida (Almeida's moat and a fausse-braye with two tiers of gunports and circular comer bastions), in Castelo Rodrigo (torre de menagem [castle keep] and barbican bastion), in Castelo Bom (comer gun-tower and other elements), and elsewhere in the Ribacoa region. Funds and priorities were unevenly distributed across the sites--witness the dilapidated state of Sabugal's walls, as well as those of Castelo Mendo and even Castelo Rodrigo, as late as 150924--but Danzilho's execution of those improvements that he managed to put in place garnered sufficient praise to earn him an invitation to tender a proposal for the ehancement of the fortifications of Ciudad Rodrigo, on the Spanish side of the border, in 1510. His estimate, which included urgent repairs, reinforcing the gates with barbican bastions, adapting the walls to the use of firearms, and compensating for dead angles in the defensive fields of fire, amounted to some 2, 400,000 maravedis. (25)

The concejo of Ciudad Rodrigo having apparently balked at the high cost of his estimate, Danzilho had no trouble securing alternative assignments on an even grander scale back in Portugal. In 1511 he was contracted to renovate the defenses of all the Portuguese outposts in northern Morocco--Ceuta, Alcacer Seguer, Tangier, and Arzila. To this end, he recuited about 800 of his compatriots, stonecutters or craftsmen who were also trained crossbowmen and were expected to serve in a dual capacity, as workers and soldiers. Brawls punctuated their southward march through Portugal, sparked off mainly by incidents involving women, which tended to mark the passage of every body of sixteenth-century troops, even in friendly territory. Having shipped out through Tavira safely but not without some heads bloodied in clashes with the locals, the Cantabrian/Basque contingent was ostensibly well received in the alem-mar. Three hundred were stationed in Arzila, three hundred in Tangier, and the remainder was split between Ceuta and Alcacer. (26)

Danzilho represented an interesting embodiment of early sixteenth century military entrepreneurship, serving not only as mestre de obras and building contractor (empreiteiro), but also as the captain of his crossbowmen. There is no denying either his military prowess, or his ability to think creatively and work out design changes on the fly. Moreover, he cut a dashing and unconventional figure, somewhat larger than life. The chronicler of Arzila, Bernardo Rodrigues, has nothing but warm admiration for the Cantabrian, whether he is portrayed cleaving through attackers atop the Baluarte (or cubelo) do Bispo in Tangier with a two-handed sword, or refusing a horse offered to him by D. Duarte de Meneses and preferring to march on foot at the head of his column of craftsmen/crossbowmen, armed with only a round buckler and his trusty two-hander, in an episode that Rodrigues infuses with socially mordant undertone. In the 1511 siege of Tangier, at the very start of his contract, Danzilho suffered multiple wounds, including a crossbow bolt through one of his arms and another one through the scrotum. Yet these inconvenient injuries did not prevent him from fighting both on foot and on horseback at Tangier the following summer, and acquitting himself rather well. (27) Popular with the rank-and-file, 'Danzilho was more than sufficiently earthy and impetuous, however, to rub the wrong way both careful bureaucrats and some highborn superiors.

Danzilho seems to have enjoyed at least temporarily the Crown's confidence, although the source of this trust remains to be explained. Unless it is largely bluster, a self-assurance partly reflecting Danzilho's belief that he had the King's ear, come what may, certainly seems to emerge from the 1514 auto de medicao. Some of his clashes with Diogo Boytac may have been sparked off both by this posture and by a gregarious disregard for the niceties of procedure. Having succeeded Boytac at Arzila as mestre de obras, Danzilho won many hearts by his liberal help to the moradores of Ajrzila in mending the damage to houses and property inflicted during the Moroccan assault of 1508. Apparently he never turned down any request for a cartload or two of mortar, and even assigned his workers to assist the inhabitants, lending the services of his stonemasons, here for a day, there for a whole week. As Bernardo Rodrigues notes, no other royal official since Danzilho's time behaved so, and he credits Danzilho with making possible the bulk of repairs to Arzila's civilian infrastructure. Yet the flavour of the passage also implies that the approach won him few friends among escrivaes, vedores, feitores, and other ranking officials. (28)

Diogo Boytac (or Boitaca) (e. 1450-c. 1528) seems to have come to Portugal from France, possibly, as some have speculated, from the Languedoc. The years 1500 to 1516 saw him at the peak of his career. His mind shaped the design and construction of the church of the Convento de Jesus in Setubal,, and then fully asserted itself in Santa Maria de Belem (Mosteiro dos Jeronimos) (1502-16). At the same time he was engaged in putting the final touches to the Se da Guarda (1504-17), and in redesigning and remodelling the monastery of Santa Cruz de Coimbra (1508-13). In the monastery of Santa Maria da Vitoria (Batalha) his craftsmanship left its mark in the Arcos do Claustro Real as well as in the Capelas Imperfeitas (1516 onward). Other works attributed to thim include the Pacos Reais of Mondego, and the churches of Freixo de Espada a Cinta, Olivenca, and Golega. Even the initial concept of the emblematic Torre de Sao Vicente de Belem (Baluarte do Restelo), built by Francisco de Arruda in 1514-1520, has been ascribed at least partially to Boytac.

Mestre Boytac became involved more closely in refurbishing the fortifications in Morocco in 1509-10, when assigned to Arzila under the heroic but also brusque and somewhat controversial governor D. Vasco Coutinho, Count of Borba, who also intermittently served as a member of the conselho real. (29) Arzila was still reeling from the effects of the furious and very nearly victorious Muslim assault of 1508, followed in 1509 by a second Wattasid siege, part of a series of counteroffensive operations that would culminate in two more attempted sieges in 1511-12. (30) Boytac's performance in directing work on the walls as well as in planning the dry ditch, besides other building tasks, earned him a dubbing to cavaleiro by the Count. In 1511, he was elevated to cavaleiro-fidalgo of the Casa Real, an advancement in which D. Vasco Coutinho's patronage appears to have played a significant part. (31)

In the late spring of 1514, Boytac returned to Morocco, essentially in the capacity of a royal building auditor. His task was to measure completed work, inspect compliance with instructions, verify the amount and quality of material, and approve relevant supplies and work as satisfactorily delivered, preliminary to payment. To some extent, this might be perceived as a mundane chore. Few other names than his were more intimately associated with the very essence of the Manueline architectural style, and the projects for which he had been responsible in Portugal included some of the showiest jewels of the early sixteenth century. To audit the Moroccan fortresses was an ambiguous assignment, a position of trust but also one that was pedestrian given Boytac's past accomplishments. By 1516, two years after his North African tour, he seems to have partly fallen out of favour with the Crown, as Paulo Pereira noted, although in accounting for his relative eclipse one certainly cannot ignore his advancing age (he was around sixty-five during the audit at Alcacer). (32)

The tensions between Danzilho and Boytac that emerge from the 1514 auto de medicao were real, but they have been somewhat exaggerated. (33) Boytac and his assistant, Sebastiao Luis, simply behaved at least in part as their role dictated--as bureaucrats mindful of the royal purse and instructions--whereas Danzilho was quite true to his unconventional self. This polarity certainly did not make relations within the trio any smoother, but the points of contention were less portentous than they might seem.

Three main sources of friction can be identified. Firstly, Danzilho showed a propensity to let himself be guided by his military judgment instead of his instructions, and delivered better work than specified--and thus more expensive--in those areas that would have mattered to a practical soldier: gunport embrasures, stronger parapets, improved gateways, additional bastions, and solidly flagstoned high-traffic areas over which guns might have to be hauled. Typically he used more ashlar than he should have if he simply followed instructions, and put dressed stone where mere brickwork was deemed sufficient by the regimentos, (34) The approach seems to be quite consistent with the tone and arguments of the proposal he had submitted to the consejo of Ciudad Rodrigo the year before shipping out to Morocco. By that time he had clearly acquired a set of fortification concepts with which he felt comfortable both as a fighting man and an engineer, and applied them in different variations but with the same underlying pattern to most of his projects.

Secondly, Danzilho skimped somewhat on masonry sections not directly exposed to gunfire, particularly the wet ditch bottom and the counterscarp, and probably reused stones from those portions of the thirteenth- to fifteenth-century walls that his workers pulled down in the process of modernizing the defences. Given the layout of Alcacer's fortifications and the construction stages discussed below, construction seems to have been expedited by partly lining the ditch with smaller stones taken in situ from an old fausse-braye on the east and south side of the town. Danzilho also commandeered whatever came to hand, even material left over by previous builders that, from an accountant's point of view, was not his to handle. (35) Thirdly, he carried out more work than originally specified on providing suitable amenities within the castelo, particularly in the area of the governor's stables and of the castle bakehouse. On the whole, he seems to have attempted to stay within a budget, but he also built to last and to withstand, and to suit soldiers like himself rather than escrivaos.

Boytac unfortunately set the tone by grumbling from the very beginning about the ashlar and flagstones in the counterscarp, and by asserting that the stonework in the ditch was of poor quality and not well seated in mortar, although how he knew this begs a question, given that when the survey began, on 12 June 1514, the ditch was full of water. Matters escalated when the old mestre further argued that the stones he saw were too small and badly dressed, more like quarry rejects than proper ashlar. Finally, he asked pointedly to see the schedule of the dressed stone Danzilho was contracted to deliver, only to discover that no one had any idea where it was. It was attached neither to the regimento, nor to any of the various supplementary contracts. That the ditch was an important sticking point is confirmed by the fact that in the end Boytac insisted on having the scarps and counterscarps opened up in twenty different test-spots, only to reach the conclusion that the work was to specification after all, and could not be faulted, at least not overtly. (36)

Danzilho also ran afoul of Boytac, at first, over extensive additional work not covered by the latter's copy of the regimentos, yet apparently executed in good faith on the authority of another regimento that Duarte Ferreira left behind in Alcacer, as well as on the basis of letters received directly from the King. (37) Danzilho finally had to request expressly that all such items be included in the auto de medicao, so that he could clear matters up and press for payment once he had returned to Lisbon. Another sore point was a quantity of miscellaneous dressed stone (477 pieces, to be precise), which Boytac summarily condemned as too small, too badly dressed, or set in places where they were not needed or where brick could very well have been used instead. Expostulating that he would go over this with the King in person, Danzilho had to settle for the stones being counted and listed, but not approved for payment. (38)

Yet in a number of instances Boytac had no trouble agreeing that where Danzilho exceeded instructions, for instance in the angled bends of the ditch, the result was indeed preferable--"para seguranca da obra." (39) Likewise, he approved of the six ashlar-framed gunport extensions in the Baluarte da Praia glacis, made necessary because the glacis rose higher up the wall than the lowermost gunports. (40) He expressed few reservations about the stonework around the Porta de Ceuta drawbridge, its supports, the adjacent scarp and counterscarp, and the jambs of the gateway. Outright praise even blessed the rear security walls, not provided for in the regimento, which Danzilho installed in order to close off the backs of the three traveses between the Porta de Ceuta and Porta de Fez (described and analyzed below). (41) Still, Boytac's willingness to overstep written instructions clearly had its limits.

A rather prominent bone of contention was the outer face of the Porta de Ceuta which, as the archaeological record suggests, still essentially preserved its Islamic features. The regimento had stressed that no changes were required in this area, but in the process of building the bastion in front of the gate Danzilho brought down a portion of the old Islamic facade ("... a qual Francisco Danzino desfez quando desfez a barreira") and remodelled the entranceway. (42) Among the lesser issues, while Boytac freely admitted that doubling the ashlar blocks in the ten gunports of the Baluarte da Porta de Ceuta was a good idea, he argued that brickwork should have predominated because it was so ordered. He also bridled at Danzilho's increasing the number of the Baluarte's parapet gunports from 10 to 14 (a principle Danzilho applied elsewhere at Alcacer, invoking need for firepower)--all framed in ashlar, not brick, in order to offset the relative thinness of the parapet. (43) The same debate arose over the gunports of the traveses, for each of which the regimento specified only four pieces of ashlar to stiffen the brickwork, whereas Danzilho defended his general use of dressed stone. While Boytac agreed that Danzilho was right, he regretfully declined to slate the supernumerary ashlar for payment, invoking the regimento, (44) On a final and minor note, where Danzilho used good masonry in the castelo's bakehouse and in the governor's stable troughs, Boytac argued that stone bound with clay would have been quite sufficient for such purposes. (45)

The sparring between Boytac and Danzilho returns us to the issue of the reliability and usefulness of the 1514 audit. Were the item counts and measurements realistic? To what extent can historians and archaeologists trust them? Matching the audit and the existing ground plans requires resolving apparent inconsistencies and potentially confusing issues of location, discussed only in the pages that follow. With the audit's key spatial points of reference identified, however, it quickly becomes obvious that the accuracy is reasonable. The audit sequence is fortunately easy to follow, given that the most significant--and thus diagnostic--perimeter measurements proceed clockwise starting from the Baluarte da Praia and end in the middle of the western wall face. An annotated table of main concordances between the audit and ground dimensions taken from the current plans is probably the most concise and accessible way of presenting the crux of the matter at a glance (Table 1). The table uses the audit's detailed linear dimensions--the right-hand "subtotals", potentially confusing to a first-time reader of the document, namely do not represent lengths but masonry volumes, calculated by Boytac and Luis to price the contractor's job.

Even though the related documents, particularly the 1502 "Instrucoes" and the 1509 "Regimento", do not offer measurements, they both partially corroborate and in turn are made more intelligible by the 1514 auto de medicao. Together, the records do not leave any significant part of the fortified perimeter uncovered, and further permit backtracking to the much sparser evidence of the fifteenth-century chronicles. (46)

Probably the most conspicuous, one might even say hallmark feature of Alcacer's late fifteenth- and early sixteenth-century enceinte was the 110 m long couraca, a sheltered fortified passageway connecting the town to the seashore and at the same time partitioning the strand into more manageable defensive sectors. The initial impetus to build the couraca must be attributed to the governor D. Duarte de Meneses, who endeavoured to assemble the basic necessary resources locally in the early spring of 1459, after the first Marinid attempt to retake the town. The bulk of the structure, however, consisting of two parallel walls, was completed with material and labour brought over from Portugal in the late spring and summer of 1459. (47)

Although improvements and additions may have been made in the 1460s and 1470s, (48) by the beginning of the sixteenth century the couraca was considered inadequate. New work was commissioned in June 1502, designed to improve the walls and equip the passageway with two round towers (cubelos) at the edge of the beach, each housing two levels of artillery casemates. The cubelos protected a stout Sea Gate, ensconced between the towers' breakwaters of dressed stone. The entire project, including the modification of the couraca battlement merlons to a broader "French" type, is well described by Dias and there is no need to dwell on the details. Dias writes, however, that the towers were to rise "... ate a altura [szcj do muro 4 palmos", (49) which makes little structural sense. According to the sources the towers simply rose, as one would expect, higher than the walls of the couraca ("... que subam sobre ha altura do muro da dita coyraca quatro palmos ...)--the only slightly suprising aspect is that the height advantage was only 4palmos or 1 m, less than a breastwork with merlons, which at Alcacer was typically about 10 palmos (2.5 m). (50)


Sometime in the late fifteenth century, a small shipyard sprang up under the couraca walls on the side facing the estuary and the bridge that crossed the Rio Canhete. This was probably little more than a medley of sheds and a makeshift slipway. In the aftermath of the 1508-9 military crisis, the shipyard was ordered destroyed, as a potential security risk. By 1511-14, certain parts of the couraca were in a parlous state, and a section nearly 47 m in length had collapsed. Francisco Danzilho rebuilt it outside of regular regimento provisions, under a special contract (asento) authorized by the governor and by Femao Lopes. According to Dias' reading of the 1514 auto de medicao, the collapsed wall was a newly built one, erected by Danzilho, who then had to correct his mistake and pay for it, but the document fails to support this interpretation. A good part of the couraca seems to have in fact gone for a long time without proper maintenance to fix wear and tear inflicted by the elements. Along the 22 m stretch of wall between the collapsed section and the castelo heavy bottom erosion and undercutting indeed required the installation of a reinforcing pediment (socalcamento). Danzilho's supposed negligence is thus much less likely to have been at fault than waves surging over the years across the shallow beach in heavy weather and shifting sand at the foot of the wall. (51)

A fausse-braye, now almost vanished and hitherto virtually ignored in the literature, was the second prominent component of Alcacer's original defenses, one that would have immediately caught the eye of a fifteenth- or early sixteenth-century soldier contemplating the town from a distance. There is perhaps no better succinct explanation of what a fausse-braye was, how it functioned, and what its merits and drawbacks were thought to be at a time when in its original form it had already become obsolete, than the relevant passage in Antoine de Ville's Fortifications, published in Lyon in 1639. (52) The pithy opening definition is worth quoting in full:
   Auant que l'Artillerie fust en usage, on se seruoit des
   Faussebrayes pour resisteraux Beliers& Catapultes, & lors on
   faisoit double muradle cede de deuant estoit plus basse, qu'on
   appelloit Faussebraye, comme en la Planche 18, Figure 1. De cette
   sorte on en voit en plusieurs lieux, comme tout au tour de la Vi
   lie d' Axles, a S. Antonin que le Roy assiegea et prit. Cette sorte
   de Faussebraye ne vaut plus rien, n'estant pas capable de resister
   au Canon. (53)

Nice, short, a propos--and describing precisely what the excavators of Alcacer found on the outside of the north-eastern quadrant of Alcacer's enceinte, ostensibly without realizing the exact nature of their discovery--a second outer wall "running parallel to the main town wall, about 5 m from it" (54) and apparently lower than the latter.

Modified fausse-brayes were still in use in De Ville's time, and some tacticians even set great store by them, as the author reminds us. These later fausse-brayes, unlike their medieval predecessors, were backed by a fill of earth and rubble forming a solid combat esplanade (the franc). (55). Rising about 8 to 10 French pieds above the bottom of a dry ditch, or 3 to 4 pieds above the water level of a wet one, with a parapet at the level of the surrounding open terrain, a fausse-braye shielded the lower part of the main fortress wall, and arguably constituted a superior firing position against attackers seeking lodgment on the ditch counterscarp or down in a dry ditch. De Ville argued that fausse-brayes were on the whole best used with wet rather than dry ditches. If a fausse-braye stood too high, however, and was damaged by the enemy, it only helped to fill in the ditch, an effect compounded by rubble from the main wall tumbling into the franc. The fausse-braye thus hastened the build-up of a rubble slope eventually scaleable by the enemy. Moreover, assailants who reached the fausse-braye were at that point too close to the wall to make a good target for the defenders' heavy guns, unless the franc was swept from the lower gunports of adjacent bastions. Finally, defenders in the fausse-braye risked being hit by debris of the main wall and its parapet dislodged by incoming shot, and De Ville thus recommended a sloping wooden roof over the franc, and a secondary narrow ditch right at the foot of the main wall in order to absorb rubble through a gap between roof and wall. Whether in its old or new incarnation, a fausse-braye oifered good forward defence, but once breached it could be more trouble than it was worth.

It is safe to say that no other part of Alcacer's ground plan has been more misunderstood and misinterpreted, and the same is true of various references to the town's fausse-braye in the Portuguese documents and chronicles. The first identification of the excavated feature in the north-west quadrant as a medieval fausse-braye was to the best of my knowledge made in an article published in 2000. (56) The Redman team, by contrast, hypothesized that the wall in question was a second couraca,
   this one extending out from the citadel along the town walls to the
   point where the wall comes close to the river. Only further
   excavations would clarify whether or not this served as an entrance
   and safe passage from the river in a fashion analogous to the
   couraca leading from the sea. (57)

Yet, even though still repeated by Dias, (58) the couraca notion is undermined by all the written sources, from Rui de Pina's Chronica d'El-Rei D. Affonso Vand Gomes Eanes de Zurara's Cronica do Conde D. Duarte de Meneses to the 1514 auto de medicao. There is little doubt that the wall in question is a segment of what contemporaries described from the first day of the 1458 siege as the barreira or muro de barreira. Most of the sources make a distinction between this baireira or its various parts and the muro da vila, i.e. between the fausse-braye and the main curtain wall. (59)

A good idea of how the barreira/fausse-braye would have looked in relation to the walls of Alcacer Seguer can be gleaned from Duarte de Armas' Livro das fortalezas--standard fausse-brayes indeed surround the enceintes of Castelo Rodrigo, Moura (on whose ground plan the fausse-braye is explicitly identified as a barreira', "esta bareyra tem d'altura a lugares 7 varas e a lugares 8 varas"), Monsaraz ("esta bareyra tem d'altura 4 varas"), Campo Maior, Arronches, Monforte ("esta bareyra tem d'altura 6 varas"), Nisa ("altura da bareyra 3 varas"), Castelo Branco, Penha Garcia ("esta bareyra tem d'altura 3 varas 1 p"), Sabugal ("esta bareyra tem d'altura 6 varas 1 p"), Eivas ("muyto booa bareyra & caua"), Lapela ("altura desta bareyra he 4 varas"), and others. (60) The dating of Alcacer's Muslim fausse-braye was remains uncertain. Fausse-brayes/barreras!barreiras had been a commonplace feature of military architecture in the Iberian Peninsula, including Muslim al-Andalus, ever since the thirteenth century. The second half of the fifteenth century, and particularly the period of the war of succession following the death of Enrique IV, witnessed a proliferation of artillery fausse-brayes in the Peninsula. As to Portuguese military architecture in the reign of D. Joao II and D. Manuel I, it has been characterized, perhaps a little glibly, as mainly a matter of "barreras, cubos, y baluartes". (61) Given that Alcacer's original fausse braye does not seem to have incorporated either projecting towers (which begin to appear more often in the early fourteenth century) or artillery casemates (a feature of the later fifteenth century), it may.have been built together with the main curtain wall in 1287-91, or added not long thereafter. The renovations undertaken at Alcacer in the first half of the fifteenth century, however, may have included some improvements to the fausse-braye. (62)

It was not the main wall but the barreira, or a portion of it adjacent to one of the town gates, that the Portuguese breached first when attacking Qasr al-saghir in 1458. Subsequently they broke through one of the barreira gates, i.e. though an "outer gate", before attempting to tackle one of the town gates proper, which proved altogether too strongly fortified to yield easily:
   ... logo derribaram um grande lanco da barreira, e os cavalleiros e
   gente do Infante D. Anrique, com muito esforco e ardideza romperam
   e entraram por as portas da mesma barreira, e foram com muita
   ousadia cometer com engenhos as portas da villa, que por sua grande
   fortaleza nao poderam quebrar ... (63)

The account of the fall of Qasr al-saghir found in Weston Cook's Hundred Years War for Morocco is thus misleading: "Guns were among the items brought ashore. The next morning the gates shattered, and Qsar as-Saghir fell on the third day." (64) There is no evidence, however, that any of the main gates were shattered by artillery, a rather spectacular event the chronicles presumably would not have failed to highlight--only the fausse-braye was laboriously penetrated by the Portuguese forces by the evening of Tuesday, 17 October 1458. It took a renewal of bombardment around midnight, which had mostly a psychological effect, followed by simultaneous multi-pronged assaults combined with negotiations, to persuade the defenders to capitulate in the small hours of the morning. Incidentally, Redman's account (based on Jay Schulman's 1978 unpublished manuscript), is also somewhat erroneous. Afonso V certainly did not march into the city "on the morning of Wednesday, 18 October 1458". (65) As Ruy de Pina attests, hostilities were suspended as soon as the defenders agreed to surrender, sometime before dawn on 18 October. The Muslim inhabitants then took from daybreak until noon to file out through the gates with their movable possessions, and only in the afternoon did the king enter the town, on foot, accompanied by the Infantes and other noblemen.

The barreira, quickly repaired, it would seem, after the damage inflicted during the initial Portuguese assault, necessarily continued to play a very important role in the days and years following the conquest. During the first Moroccan countersiege of Alcacer in the winter of 1458, as well as during the second siege of 1459, the barreira constituted the first line of defence and one of the positions that at delicate junctures were put under the command of prominent fighters and charismatic personalities. (66) In most cases the barreira served as the main jump-off point for sorties and incursions against enemy gun positions and retrenchments. (67) Portuguese cavalry also used the barreira, as well as the couraca, once the latter was built, as a screeen from behind which to launch counterattacks. (68) The bandeira was low enough for Joao Afonso, for instance, to jump down and hit the ground fighting, brandishing a lance. Similarly, when D. Afonso de Vasconcelos decided not to wait for a ladder, he let himself dangle, fully armed, off a merlon in order to jump down--unfortunately dislodging some loose stones which hit him as he reached the ground. (69) To suppose that Alcacer's barriera was much higher than 3 to 4 m (within the lower range of barreira heights attested in the Livro das fortalezas) would thus be unrealistic. As one would expect, this relatively low profile, which made it expedient to sally over the wall and quickly retreat to safety using short ladders, sometimes perilously exposed the defenders to crossbow bolts and other projectiles. (70)

The Cronica de D. Duarte de Meneses leaves no doubt that the barreira/fausse-braye encircled the better part of Muslim Qasr al-saghir's medieval enceinte, a configuration maintained during the early years of Portuguese occupation, when both time and resources to effect major changes were lacking. The existence of the fausse-braye along the northern facade of Alcacer is documented in Chapter 52, dealing with Rodrigo Afonso's sortie against a Muslim retrenchment on the praia, very probably near the area of the future couraca (which had not been built yet--the events are those of the first winter counter-siege), in Chapters 54 and 55, which describe breaches made by assailants in the barreira "da parte do mar", as well as in Chapter 59. (71) The allocations of command and watch posts to notable fidalgos in 1459, outlined in Chapter 76, show that a segment of the barreira ran in front of the wall stretching from the Porta de Fez to the Porta de Ceuta, and again from the Porta de Ceuta to the couraca (thus blending into the section that, as we have already seen, lay "da parte do mar"). (72) Together with the length of wall discovered by the Redman team, this accounts for virtually the entire fortified perimeter. (73)

If any doubt still remained that a barreira shielded for instance the entire eastern frontage of the enceinte, it should be sufficient to invoke Chapter 80 of the Cronica'.
   ... o assentamento daquella uilla he em lugar chaao cercado de
   serras e da parte de Cepta [i.e. to the east] esta huma grande
   sobida que se comeca logo acerca da barreyra e uay assy sobindo
   pera cima em grande costa em razoada alteza, a que nos em este
   nosso liuro em alguuns lugares chamamos ho outeyro das uinhas. (74)

The ground at Alcacer indeed begins to rise just east of the town walls, sloping toward the higher ground still known in 1514 as the "Vinhas". Vineyards, probably replanted several times, remained scattered throughout the area despite all sieges and conflicts: the plot relatively closest to the town was the Vinha de D. Isabel, just north-east of the Porta de Ceuta. (75) It is on this side of Alcacer, as well as along the southern frontage, that the fausse-braye was replaced, possibly as early as 1502 but more probably after 1511, (76) by the wet ditch featured in the Redman plan of Portuguese Alcacer.

Finding remnants of the southern and eastern segments of the fausse-braye might naturally be quite difficult, given the already mentioned likelihood that Francisco Danzilho reused part of the barreira stones to line the wet ditch. Furthermore, the 1514 auto de medicao makes it clear that the Baluarte da Porta de Ceuta was built in part directly on top of the old fausse-braye, made in this sector of stone bound with only clay, not mortar. Excessively soft ground, noted by Danzilho and corroborated by other officials, made it necessary to reuse the earlier wall as a foundation. The barreira segment in question was still in place and intact prior to 1511--it had to be razed to make room for the Baluarte--and jutted as far out in front of the old Islamic gate as the latter, perhaps forming a barbican or a full-fledged outer ward, a feature discussed neither in the excavation reports nor in subsequent analyses. The "third" and "fourth" faces of the Baluarte, which sat on the barreira stump (given the sequence of the 1514 measurements this means the main and longest face, containing the drawbridge gate, and the face immediately adjacent to the south-west), consequently turned out wider than originally planned. As already noted, when Danzilho razed the barreira, he also demolished portions, presumably decayed, of the old Bob al-Sabta's outer gateway and facade. With its new gunports, bartizan, and pieds d'araignee supports rising up to the merlons, the gateway (ostensibly remodelled without explicit regimento authorization) must have taken on an appearance reminiscent of Danzilho's design for the bastions of Almeida. (77)



What was, by comparison, the status of the western stretch of the batreira, the one found by the archaeologists and tentatively identified at the time as a second couracai The 1509 "Regimento das obras" attests that at the end closest to the old Sea Gate (Bab al-bahr) the fausse-braye was still intact. (78) The same was the case in 1514 at the opposite end of the western segment of the fausse-braye, which had been adapted to block and control the head of the by then completed wet ditch: a barreira homde a caua fenece" Other references in the 1514 auto de medicao ("... ate o rio homde fica a barreira ...") also confirm that the remnant of the western fausse-braye was still in use. It is furthermore likely that this stretch, kept in service the longest, was the one whose merlons were modified to emulate the broader "French" model, which the 1502 "Instrucoes" used as apattem formodernizing the couraca. (79) The 1514 auto de medicao suggests, however, that the builders ultimately intended to demolish the fausse-braye entirely or in part: "[contrachapa] ... vay entestar na barreira que ficou por derribar, junto com a couraca nova que se hora faz." (80) This planned elimination of the fausse-braye seems to run against widely acknowledged trends in late fifteenth to early sixteenth century Iberian and particularly Portuguese military architecture, which favoured the use of barreiras. Its logic, however, matches Francisco Danzilho's emerging preoccupation with fields of fire, lines of sight, and with a properly functional disposition of bastions (baluartes) and baluartejos, already palpable in his 1510 proposal to refortify Ciudad Rodrigo. (81)

Some confusion might arise at this point from the mention of a couraca nova, which of course confirms that there indeed was a "second" couraca at Alcacer. This, however, was not at all the same as the western barreira!fausse-braye just discussed. What the existing literature has to say on the subject of the "second" couraca is largely misleading. The so-called couraca nova, virtually completed by 1514, was the structure that in the Redman ground plan appears as no more than a stubby tower-like projection jutting out of the town wall in a position five towers to the south-west from the castelo and three towers to the north-west from the double-pointed Bastiao Sul in front of the Porta de Fez. Originally, back in 2000, I surmised that this little discussed feature was all that remained of a full-fledged couraca running westward toward the Rio Canhete and, misled by a partial analysis of the sources, I labelled the hypothetical component a "barrier wall". This turns out to have been an error. The entire "second" couraca/ "barrier wall" issue can now be laid to rest using conclusive evidence and corresponding measurements from the 1514 auto.

The couraca nova turns out to have been nothing more than a semi-detached bastion: "... o baluarte que esta antre a Porta da Ribeira e homde se acaba a cava, que no regimento se chama Coiraca". It jutted beyond the fausse-braye wall and toward Rio Canhete no more than 22.33 palmos (5.58 m), and was linked to the main curtain by a service wall that started at an old round tower (cubelo) and stretched 14 palmos (3.5 m) from the front of this cubelo to the rear of the bastion. The service wall was only 3 m broad, 30palmos (7.5 m) high, and equipped with breastwork and merlons on either side. Two gunports were fashioned in it so as to sweep the open space between the bandeira and the main curtain, to the north and to the south. (82) The couraca nova bastion itself was 5 m broad and packed with solid fill up a height of 5 m. This base platform supported three artillery emplacements, and the second floor, rising to a height of 9.75 m above ground, accommodated another three. The bastion was topped by breastwork and merlons, 2.5 m in height. (83) Conceptually, the couraca nova corresponded to the detached or semi-detached artillery bastions thrown ahead of walls or gateways, favoured in the Peninsula in the second half of the fifteenth century, and reflected some of the notions Danzilho had articulated in the 1510 Ciudad Rodrigo proposal. The distribution of gunports adhered to the model labelled by Cobos Guerra as "jerarquizado", with two flanking-fire ports on each level and one laying down frontal coverage. The model, used for instance at Grajal, Villalpando, and Behovia, revealed its weaknesses, at least in the Spanish context, mainly during the Franco-Spanish war of 1521, when the frontal sections pierced with gunports proved excessively vulnerable to enemy gunfire. (84) When built, however, Alcacer's couraca nova bastion reflected prevalent military wisdom.

For now it remains unclear if and when the intended destmction of the western stretch of the barreira was completed. If not, then the defensive configuration on the river side of the town would have comprised the medieval curtain wall, the barreira!fausse-braye (partially controlled by gunfire from the castelo and from the service wall of the couraca nova), and the couraca nova bastion whose guns fired towards the river as well as to the north and south along the barreira. The arrangement was acceptable, given that the river and outlying retrenchments on its far bank offered additional protection, but not really good enough. (85) A major weak point was located, as far as it is possible to determine without supplementary excavations, somewhere between the Porta de Fez and the barreira. It is here that a covered channel braced with ashlar ran from Rio Canhete to the wet ditch, bringing water to fill the latter. The 1514 auto de medicao leaves no doubt that Alcacer's ditch depended on the river, not on the sea. (86) In addition to fire from the cubelos and the curtain wall, only oblique gunfire from the couraca nova, from the head of the barreira, and probably from a traves (described further on) located just east of the Porta de Fez provided intersecting fields of fire to cover this crucial sector.

The Porta de Fez, the smallest of the three old Muslim gates, was virtually worthless as a defensive element and from the moment of the initial Portuguese conquest it represented a liability. (87) During the first two Muslim counter-sieges (1458-9), the sections of fausse-braye and wall between the Porta de Fez and the Porta de Ceuta were heavily exposed to Moroccan gunfire and attempted assaults. Subsequent steps to reinforce the Porta de Fez seem to have been rather perfunctory, however, and the 1514 auto de medicao suggests that the Porta, together with a cubelo "leaning against it", were less than structurally sound. The cubelo in question, which figures in the Redman plan of Portuguese Alcacer, appears only as a dashed-line palimpsest in the plan of the Muslim Bab al-Fas presented in Qsar es-Seghir's Fig. 3.5. (88) As the Porta de Fez was hardly suited to provide meaningful additional artillery, musket, and crossbow coverage of the approaches to the water supply channel feeding into the wet ditch, the next logical move was to close the defensive gap conclusively by building the large twocomered Bastiao Sul, which Francisco Danzilho appears to have virtuallyO completed by August 1516. (89) In combination with the couraca nova and the Porta de Ceuta, the Bastiao Sul ultimately brought the geometry of fields of fire close to an effective system, both along the Rio Canhete and along Alcacer's south-eastern facade.

The latter, however, was a line considered "mais fraca e mais perigosa" in 1458-9, and cumulative military experience of the past half-century dictated that it receive special attention. The portion of the remodelling project that Francisco Danzilho completed by the time of the 1514 auto de medicao attempted to address the defense of this extended arc of curtain wall not only by means of the wet ditch, but also by modifying three of the old round cubelos into rectangular projecting gun towers. Only one of these traveses, as the 1514 auto calls them, (90) appears in the Redman ground plan (and all plans derived from it), attached to the fifth cubelo counting clockwise from the Baluarte da Porta de Ceuta. The second traves lay closer to the Porta de Ceuta, and may have been attached to the second cubelo clockwise from the Baluarte. The third traves, finally, was located just east of the Porta de Fez, most probably either at the ninth or eighth cubelo from the Baluarte da Porta de Ceuta, i.e. the tower where the wall of the later Bastiao Sul joined the main curtain, or the next tower to the east. (91) The traveses quite naturally affected the trace of the wet ditch. Two more rectangular ditch salients identical to the one in front of the single traves documented in the Redman plan were measured in 1514, and their ashlar and flagstones accounted for. Both the missing traveses and their ditch salients must be added to the ground plan in order to represent correctly the situation in 1514.

What exactly were the specifications for the three travesesl In front of each old cubelo tower, a rectangular masonry redoubt was erected, measuring c. 7.5 by 2.5 m. (92) A security wall with an access gate closed off the back of each redoubt, facing the town wall and the old cubelo. The three outer walls were 13.5 m high, and in the "first" traves (closest to the Porta de Ceuta) they were packed solid with rubble up to a height of c. 5.25 m. The two other traveses lacked this solid core (macico), because it was not deemed indispensable. The first floor of each traves housed a casemate with three gunports, and the second floor was equipped with five gunports. Originally only three second-floor gunports in each traves were provided for by the regimentos, but Francisco Danzilho staunchly defended his decision to increase the number to five, arguing that this--the second floor--was the key gunnery position. His argument was framed in terms of firepower and past experience, suggesting at least partly a distillation of local military wisdom. The overall design, however, was very much an elaboration of the baluartejos "entre puerta y puerta" (i.e. in the middle and in front of stretches of curtain wall inadequately covered by flanking fire from existing adjacent features) that Danzilho had proposed in Ciudad Rodrigo. A preoccupation with defense articulated around multiple levels of gunports also resonates clearly enough in the 1510 proposal, and the emphasis on the firepower of the upper levels (thus favouring lighter pieces and espingards) is consistent with trends documented in Spain during the reign of the Reyes Catolicos. (93)



The traveses are a far from minor issue when it comes understanding the evolution of Alcaceris sixteenth-century fortifications. Mestre Boytac, for all his grumbling about details--as any auditor is liable to do--conceded without much struggle that the third traves not provided for in the original regimento was essential, as were the back walls (town-side) of the traveses, and that all things considered everything was very well done. There was no debating the fact that "... a vila [fica] muito forte com eles e que se foram menos que se nam poderia bem guardar o muro com a artelharia". (94) While the arc of curtain wall between the Baluarte da Praia and the Baluarte da Porta de Ceuta was well covered by intersecting fields of fire from these two works, fire from the Baluarte da Porta de Ceuta effectively controlled prior to the construction of the Bastiao Sul only the stretch of wall up to approximately the position of the second traves, leaving a large dead angle further west. The three traveses partitioned the entire area into more manageable defensive sectors, and created multiple intersecting fields of fire where the guns from the Porta de Ceuta became inffective given the curvature of the curtain wall.

The completion of the Bastiao Sul made it in principle possible to dispense with the third "unauthorized" traves, which may even have been pulled down to accommodate the new bastion. The middle traves remained intact enough to be identified by the excavators. The post-1516 fate of the first one, closest to the Porta de Ceuta, remains an open question. Given that this was apparently the one traves with a solid first-floor fill, remnants of it should have been found. It is possible that it too was pulled down to streamline gunfire geometry. Had both the first and third traves been demolished between 1516 and 1550, the defensive configuration would have arguably been more symmetrical, with fields of fire from the Baluarte da Porta de Ceuta and the Bastiao Sul intersecting neatly just in front of the remaining traves. Until now no further relevant documents have been found, and only additional excavations and wall structure soundings may eventually clarify the building sequence.

At the opposite, northern apex of Alcacer's defences, the transformation of the old Muslim Bab al-bahr (Sea Gate) into a true castelo, constituting together with the refurbished and improved fifteenth-century couraca a key tactical position, was set in motion only in 1509. Steps in this direction were being taken since 1502, but the true spur was the military conjuncture of 1508-9. Portuguese efforts to extend control over key points along the coastline of a politically fragmented southern Morocco, which gathered momentum in 1505, were followed in 1508 by the occupation of Safi and by an attempted conquest of Azemmur. Local counterttacks followed quickly, and were emulated in the Wattasid north of Morocco by operations mounted by Sultan Muhammad al-BurtuqalT, his officers, and allies. Arzila was nearly recaptured by Wattasid and allied forces in 1508, and another siege was laid to the town in 1509. Further operations against both Arzila and Tangier followed in 1511-12. (95) There is little doubt that 1508-9 represented hinge years both in terms of an escalating Portuguese commitment to war in Morocco and in terms of the ways in which Portuguese commanders evaluated Moroccan military capability. The same years constituted the starting point for the chronology of the castelo's architectural features that the Redman archaeological team hesitated to date more precisely.

The instructions given to Pero Vaz in 1509 provided first of all for the construction of a barreira on the south side (i.e. the town side) of the old Bab albahr, enclosing the lodgings of Alcacer's governor and the attached service buildings and areas (granary, storehouse, mill, baking oven) in order to make them defensible against attack from inside the town if the latter were overrun and the garrison had to make a last stand. The lesson of the 1508 siege of Arzila, which might well have fallen but for the ruthless and controversial retreat of the Count of Borba and a core of defenders into the local castelo, obviously sank in quickly. Alcacer's castle barreira started from a section of the old outer fausse-braye next to the Bab al-bahr, cut straight through the town's curtain wall in the direction of the governor's pig pen (curral dos porcos), and then turned east (roughly but not quite in the direction of the church of S. Sebastiao), toward the governor's stables (estribaria), and from there toward the arraial (muster square), located just east of the old Bab al-bahr. This barreira, with most but not all of its planned attachments, still survives and has been extensively documented both in the Redman ground plans and in Dias' photographs, even though neither adequately analyzed nor put in a proper context.

One might as well call the castle's barreira a "town-side fausse-braye"--this certainly is how the wall was conceptualized by its planners and builders, who called it a barreira in full knowledge of what they were erecting. (96) The barreira was indeed supposed to be lower than the curtain wall, and certainly lower than all other features of the castelo, so that if captured it would be exposed to fire from the inner tier of fortifications. Like a fausse-braye, it constituted the first line of the castelo" s town-side defence. At the two points where it joined the main curtain wall, two cubelos were to be erected from which the barreira could be enfiladed and which would close off access to the curtain's walkways. The cubelos were to have two levels of gun casemates, with the lower gunports permitting fire both along the barreira walls into the town, and toward the sea on the other side. The role of the upper gunports was to sweep the top of the curtain wall, as well as the barreira. The cubelos do not seem to have ever been completed, and only one original Muslim cubelo at the eastern juncture of the barreira and the curtain was partly adapted to the specified function. The sections of curtain wall between the planned cubelos and the Bab al-bahr were to be equipped with breastwork and merlons facing the town, to provide parapet firing positions against attack from that direction.

The barreira's town gate was protected by a still extant U-shaped bastion. Along the barreira, facing the interior of the town, the regimento made provision for a dry ditch, 8.8 m deep and 8.8. m wide, with a drawbridge controlling access to the gate. It is difficult to assess whether the ditch was dug as planned, given that few archaeological soundings have been carried out in the area and only one sample square was opened up by the excavators next to the south-east comer of the barreira. Structural inspection in situ suggests that a ditch, although narrower than specified, may indeed have separated the castelo from the town. The U-shaped bastion was equipped with gunports to enfilade the ditch as well as to fire into the town, across the main plaza toward the Casa da Camara and into the gullets of the two adjacent streets.

Any attempt to decipher the stmctures and the construction sequence within the castelo proper requires first of all locating a key spatial point of reference: the governor's lodgings (the aposentamento do capitam). Between them, the 1509 regimento and the 1514 auto de medicao leave little doubt that these lodgings were built directly into and on top of the former Muslim Bab al-bahr, and by 1509 had also encroached on the top battlement walkways on the town side, which were at that point ordered cleared in order to make them operational. (97) Although it is not easy to reconcile the documents and the remnants of the partly mined Bab al-bahr, it would be logical to assume that the upper floor of the aposentamento sat for the most part atop the original western entrance chamber of the city gate, above its hemispherical ceiling supported onpendentives. This tallies well with the mention, in the 1509 regimento, of the "arcos mouriscos sobre que o dito apousentamento esta fundado", within the area where the governor's mill (atafona) was located. It is tempting to identify these "arcos", ordered filled in with sections of wall in 1509, with the old arched niches and portals of the eastern, unroofed gate chamber. This phase may not be the one during which the entire eastern chamber was, as Redman described it,
   reinforced on the inside with thick stone and mortar walls in the
   form of an X reaching into the four comer. This was filled in with
   an enormous volume of loose rubble and then surfaced over,
   providing the Portuguese with a solid platform 11 m above the
   surrounding ground level ... (98)

It certainly provides, however, a plausible date for the archaeologically documented filling of the arches with "parede ... para ficarem mais fortes e da grosura das outras paredes." (99) Any ground plan sketches that link the fill to the same building phase as the core of the couraca (i.e. mid-fifteenth to later fifteenth century) are thus unlikely to be correct.

The integral filling of the second gate chamber is inseparable from the issue of the exact configuration of Alcacer's castle keep (torre de menagem). The keep constitues one of the castelo's more confusing features, mainly because the instructions relating to it are not as clear as one might wish, and seem to have changed over time. Dias has drawn attention to a report by mestre Martim Lourenco, dated 13 July 1507, which implies that the torre de menagem was under construction at that point and that it was already "cheia ate ao sobrado". (100) The location is unequivocal--"sobre a casa das atafonas", i.e. above the gate chamber with the "arcos mouriscos" mentioned earlier. Unfortunately, if the atafonas were still in place and used in 1507-9 and if the second gate chamber was thus not blocked up as yet (otherwise there would have been no acces to fill the "arcos mouriscos" with "parede" in order to bring the walls to a uniform thickness), we must necessarily wonder how the 1507 version of the torre de menagem was to be structurally supported, and how it could have been already "filled solid up to the second floor".

The 1509 regimento adds further complications. Firstly, its wording seems to suggest that the torre, in a fairly advanced stage of completion according to Martim Lourenco, had in fact not been built yet. Secondly, the square torre specified by this particular regimento was to be anchored to the walls of the couraca> which were ordered reinforced should they prove structurally unsuited to support the keep. The instructions for the keep's ground level included a core of solid masonry up to a height of 2.2 m. An upper storey drawbridge was to constitute the only point of access between the torre and the governor's lodgings. To reach the couraca from the keep, the garrison was supposed to use a doorway positioned at an appropriate height and equipped with stone supports for a ladder or staircase. The regimento also ordered the building of a traves starting at the couraca and running around the keep to join the muro of the castelo--and as the latter "muro" is said to be lower than the couraca it is tempting to interpret it as the castle's barreira (the fausse-braye discussed earlier and separating the castle from the town). A gateway pierced through the old town wall was to provide access to the couraca by going "around" the keep. The regimento then specified a second gateway in the couraca wall next to the keep, between the latter and the spot where traves and couraca joined, unless the builders deemed that the couraca between the keep and the traves should simply be demolished. (101)

While this for the most part fails to match the current configuration, whether on the ground or documented through excavations, the regimento is in fact rather straightforward, despite any confusion that it might inspire. It represents an attempt to rethink fundamentally the original ad hoc solution of 1506-7, which had been to create a torre de menagem by reusing the eastern portion of the Bab al-bahr. It would certainly have been quite feasible to erect a torre on the foundations provided by a strengthened but not blocked eastern gate chamber (casa das atafonas). Martim Lourenco may also have contemplated using merely the old eastern exterior gate tower, indeed already solid up to parapet level. The 1509 regimento, however, ordered the building, obviously never even commenced, of a new square torre de menagem right in the throat of the couraca, as wide as the latter and entirely separate from the Bab al-bahr. Access to the couraca from inside the castelo would in this case have led around the eastern side of the keep, turning west through a gateway into the couraca and then north to proceed along the latter--in effect a bent entryway. The traves stretching east and then south toward the point where the town-side barreira met the curtain wall was probably intended to enclose a space equivalent to Redman's patio d'armes and the early sixteenth century arraial. Gunports in the traves, facing the beach, were designed to cover the east side of the couraca. (102)

The 1509 solution was suprisingly antiquated, all things considered. Fire from the torre de menagem would have effectively swept the length the couraca (the regimento even specifies that the new couraca towers at water's edge should have their backs opened up, so that they would be exposed to fire from the keep if they fell into enemy hands), as well as a portion of the inner bailey enclosed by the traves. It would have also partly obstructed, however, fire into the couraca from the top of the Bob al-bahr and, depending on the keep's height, fire into the open field east and west of the couraca, both from the Bab al-bahr, from the adjacent cubelos, and from the curtain wall. Relatively ineffective as an artillery platform, a square keep aligned with the couraca would also have had its four dead angles facing to some extent in awkward directions. Finally, the entire essence of the design was oriented primarily toward defence in hand-to-hand combat and toward gradual retreat through the couraca to waiting ships, instead of toward active long-range defence using artillery.

The only portion of this design to be implemented, however, was the traves. Whether or not entirely completed, it provided the core of the subsequent L-shaped Baluarte da Praia, erected by Francisco Danzina in 1511-14. The keep in the throat of the couraca remained a still-born idea. Instead, the eastern gate chamber, used as the casa de atafonas, was closed up after 1509 and blocked with the massive cross-braced fill described by Redman, thus creating an artillery platform whose pre-existing orientation was in fact rather nicely suited to enfilade the couraca as well as to provide reasonably distributed fields of fire for guns firing toward the open ground, over the outlying defences. The earlier 1507 notion thus prevailed over instructions sent from Lisbon. Needless to say, no matter who its originator was, it represented a solution formulated unlike the 1509 regimento from sound knowledge of the place. It would also have been quite compatible with Danzina's 102 cavalier approach to overly detailed instructions issued by armchair designers, and it matched the just emerging tendency to lower a fortifications' profile, limiting the height of conspicuous such as a traditional torre de menagem, highly vulnerable to gunfire.

Where the newly completed Baluarte da Praia with its two levels of gunports joined the couraca, Danzina built a wall across the open throat of the latter, 11 m high and 11.8 m long. This is the wall whose narrow arched gate still provides access from the castelo to the present-day remnants of the couraca. On each side of the gate, two gunports were fashioned, and another two were installed in the parapet on top of the wall (a parapet that, with its merlons, brought the wall height to c. 13.5 m). Between the head of the western couraca wall and the base of the eastern exterior tower of the Bab al-bahr, a connecting service wall (corredoira) was installed at the same time, running c. 12 m long and 3.5 m high, with a 2.2 m high parapet facing toward the Rio Canbete. A staircase descended from this service wall into the second baluarte "q'esta contra a Porta da Ribeira", which can only correspond to the western outer wall of the castelo, (103) The construction date of this "bastion" is more difficult to determine. The documents project a vague impression that it was already in place at the time of the various early sixteenth-century improvements, and it may thus partly date back to the second half of the fifteenth century.

The final section of the 1514 auto de medicao that relates to the internal configuration of the castelo deals with the governor's stables, work on which was not provided for by any royal instructions. The 1509 regimento implies that the stables were at that point located well inside the old town wall, around the south-eastern comer of the town-side fausse-braye about to be put up. The 1514 auto, however, is entirely inconsistent with the stables having remained there. They would presumably have had to occupy the south-east comer of what Redman ' s team had labelled the "interior court", where a half-mined flight of stairs still leads up to what may have been the upper storey of a vanished structure and from there probably to the barreira's parapet. (104) Yet the auto mentions a wall that runs "above" or "across the top of' of the stables, joining the service wall (corredoira) that provides access to the couraca. Only one such service wall is consistently mentioned, and that is the corredoira discussed above. The stables are certainly rather difficult to visualize from the auto, but they were unquestionably moved between 1509 and 1514, and rebuilt somewhere around the arraial ward adjoining the Baluarte da Praia.

Within the area of the old arraial and of the Baluarte da Praia's inner ward, or for that matter anywhere else between the curtain wall and the couraca there is little space to accommodate the 25.8 m wall "above the stables" measured in 1514, as well as the troughs, which took up 18.7 m. One plausible spot lies right along the incompletely excavated wall separating the bastion ward (Redman'spatio d'armes) and the old arraial, and with a wall "above the stables" extending right across the mouth of the couraca to join the corredoira the distance would be just about right. As no more than two or three test squares have been opened up by the Redman team inside the castelo, it is impossible to be more conclusive. The stables had a mortared flat terrace roof (eirado), through which a masonry chimney rose, and part of the terrace space was taken up by at least one dwelling. A wooden hoarding supported on ashlar quoins ran part of the length of the wall "above the stables", and one or more brickwork bartizans (guaritas) controlled the edge of the terrace, on the side facing the arraial.

The original muster square (arraial) probably accommodated one or two small houses, one of them a two-storey structure whose second floor was extensively renovated in 1511-14. One of these dwellings, with a flat terrace roof, housed the governor's stable-groom. The old cubelo at the juncture of the Baluarte da Praia and the curtain wall, indeed one of the two cubelos supposed to anchor in 1509 the defence of the town-side fausse-braye, was transformed into a dovecote. (105) Under the walls of the old Bab al-bahr, most probably next to the stables, was a chicken run (terreiro de galinhas), and the arraial also held a small garden adjacent to the wall of a small retrenchment (tranqueira) "behind" the stable troughs. On the west side of the Bab al-bahr, most likely between the western outer tower and the juncture of the western fausse-braye with the castle wall, a partition 3 m long and 4.5 m high, with a gateway, was built in order to hold livestock. The resulting encolsure, roughly 15 m by 5 m, would have been sufficient to hold a minimum of fresh meat on the hoof. (106) Making allowance for the poultry, pigeons, stock, and vegetables within the castelo naturally served the governor's table, as well as constituting an iron reserve for military emergencies. To some extent, however, it must also have compensated for the loss of earlier ad hoc facilities (curral dos porcos, estribaria) located within the town walls, between the old Bab al-bahr, the small cemetery at the north-west end of the church (igreja matriz) of Santa Maria da Misericordia, and the Casa da Camara. The 1509 barreira and its ditch now occupied much of this space, and it remains unclear which of the castelo's service areas were accommodated within the forecourt (Redman's "interior court") enclosed by the barreira.

Seen as a whole, the fortification elements that Francisco Danzilbo grafted onto the walls of medieval Portuguese and Islamic Alcacer and that mestre Boytac measured in the summer of 1514 represent a patchwork of architectural notions and military solutions, but the ensemble fits the era, the circumstances, and the builder. The additions transformed pre-1500 Alcacer's enceinte into a transitional artillery fortification, but none of the features were really ahead of the times and some remained quite archaic (and would have seemed even more so had the 1509 plans for the torre de menagem been implemented). Alcacer, like many of Portugals' land frontier fortalezas sketched by Duarte de Armas, remained close to a minimalist solution--the medieval town wall wore only a veneer of artillery-age features, elevated above the level of a token mainly by the combined use of forward works, a wet ditch, and more or less systematized fields of fire. The town's face-lift was a one-shot affair, quite reminiscent of the managerial strategic change paradigm known as "unfreeze-change-refreeze"--a set of concepts aiming to explain why shifting a well-established business from one relatively steady-state structure and modus operand'i to another tends to involve fundamental restructuring only at infrequent intervals. (107)

Whether in the modern business world or in the business of sixteenth-century military expenditure, the factors that favour working routine stability and thus only fitful change, as opposed to a relentless but ultimately draining search for an optimum are likely to include force of habit, entrenched frames of reference, localized vested interests, and the cost-benefit balance of being wholly "up-to-date" or of reaching the "ultimate" and invariably illusory solution, especially in view of competing commitments. To these it is necessary to add, in the context of early modern western European state finance, political will and institutional tradeoffs required to mobilize, even over a short span of time, the necessary fiscal and other financial resources. Even if the Crown in Manueline Portugal truly did have somewhat more elbow-room to maneuver financially than some of its contemporaries, it laboured under a heavy burden imposed by far-flung involvements, from Morocco and sub-Saharan Africa to the Indian Ocean, as well as by obligations to the nobility and other stake-holders within Portugal, stemming from expectations of entitlement partly rooted in tradition and partly representing the cost of securing loyalty, factional support, political consensus, institutional balance of power, and perceived prestige. (108) Penny-pinching remained a constant shadow companion of an ostentatiously advertised Manueline power and opulence. Prioritizing, sometimes sensible and sometimes inescapably corrupt, prevailed throughout the burgeoning Manueline bureacracy. Even during the 1508-18 period of military crisis and expansionist drive in Morocco, Alcacer, decidedly less threatened than Arzila or Tangier, never became a prime concern capable of firing up enthusiasm and eliciting a symbolically suitable show of dogged support. Significant because it fitted into an old-established strategic equation, humdrum Alcacer, hardly at the forefront of chronicled events, received what seemed needful, but no more.

Danzilho's profile was eminently congruent with this context: his documented building career was mainly that of a hands-on refurbisher, and it would not be at all surprising to discover that he promoted himself as one who could "fix up" decrepit fortifications quickly and for a reasonable sum. This is not meant to suggest that men like Danzilho were military technology charlatans (although the sixteenth century did have its fair share of those). The tone of the documents suggests that Danzilho's approach resembled that of the contemporary Spanish engineers mentioned for instance by Cobos Guerra, from Ramiro Lopez and Diego de Vera to Pedro Malpaso and Pedro Navarro, many of them artillerymen who had learned their design trade both facing enemy batteries and behind the sights of their own guns. (109) From his days in the Ribacoa, through his invitation to refortify Ciudad Rodrigo, to his Moroccan assignment, Danzilho clearly believed that a judicious use of well placed artillery barreiras, projecting gun towers, gate-shielding bastions, and other adaptations could give old walls a new lease on life. At that point in time, he was more or less right--and his way had the merit of being affordable. The approach certainly presaged Machiavelli's often cited pragmatic option for modernizing existing walls, set out in a report submitted to the government of Florence in 1526. (110)

Circumstances had changed significantly, however, by 1549, when D. Pedro and D. Joao de Mascarenhas, seconded by Miguel de Arruda and Diogo Telles, finally threw up their arms and recommended abandoning Alcacer on the grounds that it could not be rendered defensible. (111) Between 1520 and 1550, the new "artillery fortress" with an increasingly intricate geometrical arrangement of components based on the so-called bastioned or "Italian" trace had begun its inexorable proliferation across Europe. In Portugal, high-level political and fiscal will to sustain an assertive presence in Morocco had been waning since the 1530s, amid factional maneuvers and wrenching political debates at the court of D. Joao III about the relative merits of continued involvement in North Africa, a permanent shift of focus to the Indian Ocean, or even wholesale colonial divestment. (112) In 1541, the commitment had been made to transform Mazagan into Portugafs showpiece artillery fortress in Morocco, partly because here no complications arose from having to deal with an old fortified enceinte. By 1542 Mazagan was virtually completed, and Alcacer was living on borrowed time. The effort and resources spent in the spring of 1550 on fortifying Monte Seinel, (113) the top of the hill on the west side of the Rio Canhete estuary, which dominates the modern village of Qsar es-Seghir, might have been nearly enough to begin transforming Alcacer, in its new location across the river, into Portugal's second modern fort in Morocco, but the symbolic significance and obvious maritime advantages of the ports of Ceuta and Tangier ultimately took precedence.

Martin Malcolm Elbl

Trent University

(1) See e.g. Charles L. Redman, Ronald D. Anzalone and Patricia E. Rubertone, "Medieval Archaeology at Qsar es-Seghir, Morocco," Journal of Field Archaeology 6 (1979): 1-16; Charles L. Redman, "Description and Inference with the Late Medieval Pottery from Qsar es-Seghir, Morocco," Medieval Ceramics 3 (1979) : 66-79; Charles L. Redman and James L. Boone, "Qsar es-Seghir (Alcacer Ceguer): A 15th and 16th Century Portuguese Colony in North Africa," Studia 41-42 (1979): 5-50; Janet H. Tewksbury, "Skeletal Remains and Socio-Economic Implications of Mortuary Practices at Qsar Es-Seghir, Morocco," M. A. Thesis (University of Manitoba, 1979); Charles L. Redman, "Late Medieval Ceramics from Qsar es-Seghir," in La ceramique medievale en Mediterranee occidentale--Xe-XVe siecles. Valbonne, 1978 (Paris: CNRS, 1980); Martin Zierman, "Qsar es-Seghir. Eine islamische Staotbefestigung des 13. Jh. in Marokko," in Burgen und Schlosser, Zeilschrift der deuischen Burgenvereinigung, e.V. fur Burgenhmde und Denhnalpflege [Braubach/Rhein] (1981): 117-124; James L. Boone, "Majolica Escudillas of the 15th and 16th Centuries: A Typological Analysis of 55 Examples from Qsar es-Seghir," Historical Archaeology 18 (1)(1984): 76-86; Charles L. Redman, Qsar es-Seghir: An Archaeological View of Medieval Life (Orlando: Academic Press, 1986); J. Emlen Myers and M. J. Blackman, "Conical Plates of the Hispano-Moresque Tradition from Islamic Qsar es-Seghir: Petrographic and Chemical Analyses," in La ceramica medievale nel Mediterraneo Occidentale. Atti del III Congresso Internazionale oiganizzato dal Dipartimento di A rcheologia e Storia delle Ani deli'Universita di degli Studi di Siena e dal Museo delle Ceramiche di Faenza (Florence: Al Insegna del Giglio, 1986): 55-68.

(2) Redman, Qsar, 14-19.

(3) Redman, Qsar, 12.

(4) Moreover, Redman's comments imply difficulties with transcribing and dating tombstones and other Portuguese epigraphic material recovered from the site and suggest certain workflow bottlenecks in the area of Portuguese palaeography. For instance, the tombstone depicted in Fig. 5.10 (Redman, Qsar, 153), is described as one of a batch "still being studied. Two of the stones begin with the words AGUY YAS ("here lies"), but the name and dates are not yet known" (Redman, Qsar, 152). Even from my bad photocopy of the page in question, however, it is readily apparent that the stone commemorates Afonso Alvarez de Olanda and is dated 1519. Palaeographic and linguistic obstacles may well have been one of the factors limiting the team's capacity to exploit Portuguese archival sources.

(5) Vieira Guimaraes, Marrocos e tres Mestres da Ordem de Cris to, Comemoracao do Quinto Centenario da Tomada de Ceuta, 3a. ser, Memorias (Lisbon: Comissao dos Centenarios de Ceuta e Albuquerque, 1916), 206-19; As Gavetas da Torre do Tombo, vol. 5, Gav. XV, Macos 16-24 (Lisbon: Centro de Estudos Historicos Ultramarinos, 1965).

(6) Redman, Qsar, 146.

(7) Weston F. Cook, The Hundred Years War for Morocco. Gunpowder and the Military Revolution in the Early Modern Muslim World (Boulder, CO: Westview Press, 1994), 137-63; Antoniao Dias Farinha, Os Portugueses em Marrocos (Lisbon: Instituto Camoes, 1999), 24-6; Martin M. Elbl, "Portuguese Urban Fortifications in Morocco: Borrowing, Adaptation, and Innovation along a Military Frontier," in James D. Tracy, ed., City Walls. The Urban Enceinte in Global Perspective (Cambridge: Cambridge University Press, 2000), 366-9.

(8) Rafael Moreira, "A arquitectua do Renascimento no sul de Portugal. A encomenda regia entre o Moderno e o Romano," Doctoral dissertation (Lisbon, 1991), 2: 24-35.

(9) My earlier reconstruction of the Alcacer fortifications, superseded by the material contained in the present article, was published in Elbl, "Portuguese Urban Fortifications," 363 (Figure 12.4).

(10) Pedro Dias, A Arquitectura dos Portugueses em Marrocos, 1415-1769 (Lisbon: Livraria Minerva Editora, 2000; reprinted 2002).

(11) Dias, Arquitectura, 60.

(12) Dias, Arquitectura, 59.

(13) Dias, Arquitectura, 62.

(14) Boytac's name is variously rendered as Boitac, Boitaca, Boytaca in the sources and their transcriptions. Most recent authors have converged on the orthography "Boytac".

(15) AN/TT, Corpo Cronologico (further CC), parte II, maco 15, doc. 84 (further in standard abbreviated format [11-15-84]) (16 Oct. 1508); Dias, Arquitectura, 100.

(16) AN/TT, CC II-9-62 (11 Mar. 1505), AN/TT, CC 11-9-152 (8 Jul. 1505).

(17) E.g. AN/TT, CCII-3-12, 11-3-44, CCII-3-48, CCII-3-53, CCII-5-48 through 52; CC IT-15-93 (23 Oct. 1508).

(18) AN/TT, CC II-9-147 (25 Jun. 1505)

(19) E.g. AN/TT, CC 11-3-52 (10 Dec. 1500).

(20) AN/TT, CCII-6-143.

(21) AN/TT, CC 11-18-131 (25 Sep. 1509).

(22) Damiao de Gois, Cronica do felicissimo rei D. Manvel, ed. by Joaquim Martins Teixeira de Carvalho and David Lopes (Coimbra: Imprensa da Universidade, 1926), Parte 3, Cap. 36: "... entre hos quaes era Francisco de lanzinha Biscainho, mestre das obras que se entao alii [in Tangier] faziao ...."; Jose Javier de Castro Fernandez," Reformas y adaptaciones de las fortificaciones medievales de Ciudad Rodrigo al uso de la artiHeria," in Isabel Cristina Ferreira Fernandes, ed., Mil anos de fortificacoes na Peninsula Iberica e no Magreb (500-1500) (Lisbon: Edicoes Colibri/Camara Municipal de Palmeia, 2002), 930.

(23) Fora definition of the term "fausse-braye", see p. 50 below (also falsa braga, falsabraga, banera in Spanish, baireira in Portuguese, and by a rather free transposition also barbacana in some Portuguese works; see Fernando Cobos Guerra, "Artilleria y fortificacion iberica de transicion en tomo a 1500," in Ferreira Fernandes, ed., Fortificacoes, 678).

(24) For a listing and brief analysis of Danzilho's (de Arcillo's) assignments, see Castro Fernandez, "Reformas y adaptaciones," 930. For the state of various fortified localities, c. 1509, see Duarte de Armas, Livro das fortalezas. Facsimile do MS 159 da Casa Forte do Arquivo Nacional da Torre do Tombo, ed. by Manuel da Silva Castelo Branco (Lisbon: AN/TT and Edicoes Inapa, 1997). The work in Almeida had been largely completed by 9 September 1508, when D. Manuel I ordered Mateus Fernandes and the stonemason (mesire pedreiro) Alvaro Pires to inspect Danzilho's work; from there they were to proceed to Castelo Rodrigo in order to assess the need for repairs and improvements.

(25) Castro Fernandez, "Reformas y adaptaciones," 930.

(26) Dias, Arquitectura, 81.

(27) Bernardo Rodrigues, Arais de Arzila. Cronica inedita do seculo XVI, Vol. 1, 1508-1525, ed. by David Lopes (Lisbon: Academia das Scienciasde Lisboa, 1915), Caps. 3-5; Gois, Manuel, Parte 3, Cap, 36.

(28) Rodrigues, Anais de Arzila, Caps. 3-5.

(29) Borba was effective governor of Arzila in 1490-5 and again in 1497-1514, with much of the period 1495-6 spent in Portugal, while D. Rodrigo Coutinho (killed in action) and D. Joao de Meneses served as interim governors. A forceful but authoritarian and acquisitive personality, Borba was denounced in 1499 for his destabilizing exploitative and heavy-handed policies toward the Muslim rural population in the hinterland. The denunciation, made by the town's escrivao dos contos not for any humanitarian reasons but to explain what amounted to a virtual appropriation of Crown revenues, throttling the payment of local tribute to the Crown, is likely to be substantially accurate (see Dias Farinha, "Portugal e Marrocos no seculo XV," Doctoral thesis [Lisbon, 1990], vol. 2, "Colectanea documental," 363-79 (doc. 167) [transcription of AN/TT, CC III-1-18]). Nine years later, Borba was largely responsible, despite his injuries, for saving Arzila during the assault of 1508, but at the cost of a ruthless command decision. To prevent Moroccan fighters from slipping inside the costeio amid the chaotic press of Portuguese refugees trying to escape from the town, which had already fallen, he ordered closing the gates while many civilians, men, women and children were still trying to reach safe haven in the castelo (Gois, Manuel, Parte 2, Cap. 28).

(30) Gois, Manuel, Parte 3, Cap. 36; Weston F. Cook, The Hundred Years War for Morocco. Gunpowder and the Militaiy Revolution in the Early Modern Muslim World (Boulder, CO: Westview Press, 1994), 145-6.

(31) Dias, Arquitectura, 98.

(32) Jose Mattoso, ed., Historia de Portugal, vol. 3, No alvorecer da modernidade (1480-1620) (Lisbon: Editorial Estampa, 1993), 435.

(33) Dias, Arquitectura, 60-1, 62-3.

(34) E.g. AN/TT, NA 769, fo. 14v. Danzilho's vocal and even stubborn preference for stonework is quite interesting, given the contemporary tendency, documented both in Italy and in various Spanish contexts (thus at Mota and Coca) to use brick in fortifications, on a significant scale. See e.g. Cobos Guerra, "Artilleria y fortificacion," 681. The trend toward brick is incidentally reflected in the Portuguese Crown's regimentos for the walls of Alcacer, both those that have come down to us and those whose instructions we know only indirectly, through allusions preserved for instance in the 1514 auto de medicao.

(35) For instance 20 apparently large and well-dressed stones left by Martim Lourenco, reused for the gate of the Porta de Ceuta, which were discounted from Dartzilho's schedule of delivered stones, AN/TT, NA 769, fo. 15.

(36) AN/TT, NA 769, fos. 8v, 9v-9bis, 14v-15, 24-24v.

(37) AN/TT, NA 769, fo. 27.

(38) AN/TT, NA 769, fo. 35v.

(39) AN/TT, NA 769, fos. 10v, 11-11 v.

(40) AN/TT, NA 769, fo. 11; the configuration of glacis and gunports is reflected in and corroborated by the Redman plans.

(41) AN/TT, NA 769, fos. 16v, 19v, 20v.

(42) AN/TT, NA 769, fo. 15.

(43) AN/TT, NA 769, fos. 15v-16.

(44) AN/TT, NA 769, fos. I7-I7v.

(45) AN/TT, NA 769, fos. 29, 32v.

(46) In this the 1514 audit and associated documents provide a nice Portuguese parallel to the Castillian Hacienda documents highlighted by Fernanod Cobos Guerra, which have likewise been found accurate enough to permit establishing close matches between the works as measured and paid for and surviving features or those documented in plans and drawings from different periods (see Cobos Guerra, "Artilleria y fortificacion," 677-8).

(47) Dias, Arquitectura, 56 attributes the impetus mainly to the king, D. Afonso V, but local initiative appears to have decidedly provided an early lead. It is useful to compare, in this respect, the two main chronicled versions of events, Rui de Pina, Chronica de El-Rei D. Affonso V (Lisbon: Escriptorio, 1902), Cap. 146, and Gomes Eanes de Zurara, Cronica do Conde D. Duarte de Meneses, ed. by Larry King (Lisbon: Universidade Nova de Lisboa, 1978), 190, 200,

(48) Dias, Arquitectura, 56.

(49) Dias, Arquitectura, 58.

(50) Gavetas, 5: 213.

(51) AN/TT, NA 769, fos. 25-26. We should not forget that tide levels at Alcacer, according to modern nautical charts, can reach two meters.

(52) Antoine de Ville Tholosain, Les Fortifications, avec l'ataque el la defence [sic] des places (Lyon: Irenee Barlet, 1639), Livre I, Part II, 123-6 (Chapitre 38, "Des Faussebrayes"). For a long time a rare print not commonly accessible, Les Fortifications has been generously made available for public download by the Bibliotheque Nationale de France on its Gallica site (http://gallica.bnf.ff), together with a number of other early fortification treatises.

(53) Ville, Fortifications, 123.

(54) Redman, Qsar, 145.

(55) Artillery casemates accommodated within the structural mass of the fausse-braye became commonplace, as a transitional feature, from the late fifteenth century onward.

(56) Elbl, "Portuguese Urban Fortifications," 363.

(57) Redman, Qsar, 145.

(58) "Uma segunda couraca deve tambem ter sido construida nesta altura, mas na direccao do rio, para Oeste, como parecem indicar vestigios encontrados, nomeadamente um duplo muro" (Dias, Arquitectura, 59).

(59) The distinction is very explicit for instance in Zurara, Duarte de Meneses, 202-3: "... e com os prymeyros tyros derribaram huum pedaco do peitorill da barreyra. E per semelhante fezerom no muro que derribarom huma amea com huum pequeno do peitoril," and again Zurara, Duarte de Meneses, 209: "... mandou chamar aquclles fidalgos e outros muytos dos que estauam fora dos muros na barreyra nas guardas."

(60) Duarte de Armas, Livro das fortalezas, fos. 25-26, 122-J22v, 124-126, 127-127v, 133.

(61) Cobos Guerra, "Artilleria y fortificacion," 681.

(62) The commemorative plaque installed sometime between 1415 and 1458 inside the first gate chamber of the renovated and reinforced Bab al-Fas (Fez Gate) certainly attests to an effort to bring the town's fortifications at least partially up to date. The Portuguese chronicles dealing with the 1458 conquest and its immediate aftermath do not project any impression of decay or disrepair where the barreira is concerned. On the contrary, the fortifications of late Muslim Qasr al-saghir, such as they were, seem to have been in good repair.

(63) Pina, Manuel, 148.

(64) Cook, Morocco, 88.

(65) Redman, Qsar, 36.

(66) Zurara, Duarte de Meneses, 160, 205-9.

(67) E.g. Zurara, Duarte de Meneses, 206.

(68) Pina, Manuel, 158; Zurara, Duarte de Meneses, 224.

(69) Zurara, Duarte de Meneses, 156, 218.

(70) E.g. Zurara, Duarte de Meneses, 208.

(71) Zurara, Duarte de Meneses, 155-6, 160-1, 168-70

(72) Zurara, Duarte de Meneses, 205-7.

(73) Zurara, Duarte de Meneses, 205-8.

(74) Zurara, Duarte de Meneses, 217.

(75) AN/TT, NA 769, fos. 27-27v.

(76) The 1502 "Instrucoes" (Gavetas, 5:214) still speak of the "barreira da dita villa" as if it were an integral complement of the "muro", and very much in existence.

(77) AN/TT, NA 769, fos. 14v-15; Duarte de Armas, Livro das fortalezas, fos. 72v-76.

(78) AN/TT, CC 11-19-106, "Regimento das obras de Alcacer," fo. 1 ("... comecais e hu traves que a bareira da vila faz junto co a porta que della say ...").

(79) Gavetas, 5:214.

(80) AN/TT, NA 769, fos. 7, 10

(81) Relevant extracts from Danzilho's proposal are reproduced in Castro Fernandez, "Reformas y adaptaciones," 930.

(82) Given its flanking fire role, the service wall might well be considered a form of traves. See e.g. Cobos Guerra, "Artilleria y fortificacion," 690, for yet another use of the term "traves". The traveses discussed below were by contrast rectangular tower-redoubts.

(83) AN/TT, NA 769.

(84) E.g. La Mota (1476-83), Ramiro Lopez' two-storey bastion of Siete Suelos in Alhambra (Granada, 1492-5), the cubete artillen of Montalban (Toledo, c. 1500), the forward bastions at Arcos or Trujillo (Extremadura), or the Baluarte Este at Salsas (1497-1503). A handy reference corpus of plans and photographs is available in Luis de Mora-Figueroa, "Transformaciones artilleras en la fortificacion tardomedieval espanola," in Ferreira Fernandes, ed., Fortificacoes, 651-7; Edward Cooper, "Desarollo dela fortificacion tardomedieval espanola," in Ferreira Fernandes, ea., Fortificacoes, 667-76; and Cobos Guerra, "Artilleria y fortificacion," 677-93.

(85) AN/TT, CC-II-3-48 (12 Oct, 1500).

(86) AN/TT, NA 769, fo. 10v.

(87) A hint of this emerges both From the crude postern gate adaptations carried out in this area, probably early on during the Portuguese occupation, and documented in Redman, Qsar, 57-58, and from the 1514 auto de medicao.

(88) Redman, Qsar, 53; AN/TT, NA 769, "... hum cubelo que esta apegado com ella."

(89) For the date, not the defence analysis, sec Dias, Arquitectura, 63.

(90) See note 82 above for varying uses of the term "traves". These three elements are not properly identified in Dias' Arquitectura, and the only passage where traveses are alluded to in connection with the 1514 auto de medicao is bound to confuse readers unfamiliar with the actual wording of the document. Dias writes of "... dispositivos que entretanto desapareceram, como os traveses que ligavam a muralha ao rio onde varavam os barcos, pois chamam-lhe da ribeira" (Arquitectura, 62). No such structures are, however, specified in the document and they indeed never existed. The only traveses measured in 1514 are the ones discussed below, not walls extending from the main curtain to the river.

(91) In any case, the traves in question was attached to one of the cubetos just east of the Porta de Fez, for the entirely methodical sequence of measurements in the 1514 auto, proceeding from east to west in this area, covers the troves' ditch salient before proceeding to give the specifications for the scarp and counterscarp in front of the Porta de Fez. The reader must keep in mind that at (his point the doublepointed Bastiao Sul was not in place yet.

(92) AN/TT, NA 769, fo. 35v. This traves "alargua mais que ho cubelo com que he encorporado."

(93) Cobos Guerra, "Artilleria y fortificacion," 681.

(94) AN/TT, NA 769, fo. 20 v.

(95) Rodrigues, Anais de Anila, 11-27 (Caps. 3-5), 33-36 (Cap. 7), 75-83 (Caps. 22-24).

(96) It notionally divided the urban space into an outer bailey corresponding to the "town", and an inner bailey corresponding to the castelo, in the same way as, on a much smaller scale, the retraimiento at Salsas cut the fortress in two. See e.g. Cobos Guerra, "Artilleria y fortificacion," 682, 685, 688.

(97) AN/TT, CC 11-19-106, "Regimento das obras de Alcacer," fo. 2v.

(98) Redman, Qsar, 146.

(99) AN/TT, CC II-J 9-106, "Regimento das obras de Alcacer," fo. 2v.

(100) Dias, Arquitectura, 60.

(101) AN/TT, CC 11-19-106, "Regimento das obras de Alcacer".

(102) AN/TT, CC 11-19-106, "Regimento das obras de Alcacer".

(103) AN/TT, NA 769, fos. 27-27v, 28-28v.

(104) Dias, Arquitectura, photograph on p. 61.

(105) AN/TT, NA 769, fos. 32-32v, 35.

(106) AN/TT, NA 769, fos. 30, 34v.

(107) Charlene L. Nicholls-Nixon, Arnold C. Cooper, and Carolyn Y. Woo, "Strategic Experimentation: Understanding Change and Performance in New Ventures," Journal of Business Venturing 15 (2000): 495; C. Gersick, "Pacing Strategic Change," Academy of Management Review 37 (1) (1994): 9-45.

(108) For examples of annual budget drafts (orcamentos) see AN/TT, NA 532 (pertaining to 1511) and 590 (1525). Both the 1511 and 1525 orcamento allocated approximately 12,000,000 reis to cover the annual expenses associated with the Moroccan outposts., covering them largely from West African revenues (AN/TT, NA 532, fo. 190; NA 590, fo. 100). For overviews of Portuguese late medieval and sixteenth-century public finances see Magalhaes Godinho's seminal study ""Financas publicas e estrutura do Estado," in Vitorino Magalhaes Godinho, Ensaios, II, Sobre historia de Portugal (Lisbon: Livraria Sa da Costa, 1968). For recent works on the 1520s orcamentos, see the relevant publications of Joao Cordeiro Pereira. For the discussion of Manueline expenditures related to the nobility, see Ivana Elbl, "The Overseas Expansion and Social Mobility in the Age of Vasco de Gama," Portuguese Studies Review 6(2) (Fall-Winter, 1998): 53-80; Susannah Humble, "Prestige, Ideology and Social Politics: The Place of the Portuguese Overseas Expansion in the Policies of D. Manuel (1495-1521)," Itinerario: Journal of the European Overseas Expansion 40 (2000): 21-45; and Susannah Humble Ferreira, "The Cost of Majesty: Financial Reform and the Development of the Royal Court at the Turn of the Sixteenth Century," in Lawrin Armstrong, Ivana Elbl and Martin Elbl, eds., Money, Markets and Trade in Late Medieval Europe: Essays in Honour of John H. A. Munro (Leiden: Brill, forthcoming 2006).

(109) Cobos Guerra, "Artilleria y fortificacion," 677.

(110) Niccolo Machiavelli, "Relacione di una vista fatta per fortificare Firenze," in Niccolo Machiavelli: Arte della guerrae scrittipolitici minori, ed. S. Bertelli (Milan: Feltrinelli, 1961), 295.

(111) For brief background and analysis, see Dias, Arquitectura, 64-65.

(112) Elbl, "Portuguese Urban Fortifications," 380-1; Otilia Rodrigues Fontoura, Portugal am Marrocos na epoca de D. Joao 111: Abandono ou permanencia? (Coimbra: Centro de Estudos de Historia do Atlantico, 1998).

(113) Dias, Arquteclura, 64-5.
Table 1. Reliability of Linear Measurements, 1514 Auto
de Medicao (AN/TT, Nucleo Antigo 769, "Alcacer")

1514                      Ground Plan (meters)   Difference
(orig. units/meters)                             (meters)

Ditch Scarp

Baluarte da Praia

21 bracas/46.2 m          c. 46.75 m             0.55 m

Baluarte da Praia to Baluarte da Porta de Ceuta

73.5 bracas/151.7 m       c. 160.3 m             1.4 m

Baluarte da Porta de Ceuta (four panes, NNE, EES, NNE, EES)

44.75 bracas/98.45 m      c. 110.33 m            11.88

Baluarte da Porta de Ceuta to Western End of Wet Ditch

84.5 bracas and 4.5       c. 195.58 m            8.55 m
palmos/187.025 m

Second Traves

10 bracas 7               22 m x 22 m x 22 m     1.75 m
palmos/23.75 m x                                 (each face)
10 bracas 7
palmos/23.75 m x
10 bracas 7
palmos/23.75 m

Baluarte da Porta de Ceuta (four panes, NNE, EES, NNE, EES)

77.5 palmos/19.375 mx     19.57 mx 11.42 m x     0.195 m
41.25 palmos/10.31 m x    39.47 m x 23.73 m      1.11 m
17 bracas/37.4 m x                               2.07 m
94 palmos/23 m                                   0.73

Second Traves

32 palmos/8 m x           8.5 m x 2.8 m x        0.5 m
10 palmos/2.5 m X         8.6 m                  0.3 m
32 palmos/8 m                                    0.6 m

Gate Wall Across Inner Throat of Couraca

47.5 palmos/11.87 m       11.3 m                 0.57 m

1514                      Notes
(orig. units/meters)

Ditch Scarp

Baluarte da Praia

21 bracas/46.2 m          E-W and N-S transit
                          measured down the
                          middle of the ditch

Baluarte da Praia to Baluarte da Porta de Ceuta

73.5 bracas/151.7 m       7 cubelos in this
                          section, cubelo at
                          junction of Baluarte
                          and curtain no counted

Baluarte da Porta de Ceuta (four panes, NNE, EES, NNE, EES)

44.75 bracas/98.45 m      Size of deviation likely
                          reflects differences in
                          assumed start and end
                          points for each wall pane

Baluarte da Porta de Ceuta to Western End of Wet Ditch

84.5 bracas and 4.5       Size of deviation
palmos/187.025 m          due to old chapas being
                          obscured by the Bastiao Sul

Second Traves

10 bracas 7               Only traves
palmos/23.75 m x          recorded in extant
10 bracas 7               ground plans
palmos/23.75 m x
10 bracas 7
palmos/23.75 m

Baluarte da Porta de Ceuta (four panes, NNE, EES, NNE, EES)

77.5 palmos/19.375 mx
41.25 palmos/10.31 m x
17 bracas/37.4 m x
94 palmos/23 m

Second Traves

32 palmos/8 m x           Only traves
10 palmos/2.5 m X         recorded in extant
32 palmos/8 m             ground plans

Gate Wall Across Inner Throat of Couraca

47.5 palmos/11.87 m
COPYRIGHT 2004 Baywolf Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Elbl, Martin Malcolm
Publication:Portuguese Studies Review
Date:Jan 1, 2004
Previous Article:Reluctant hosts and tenacious guests: Lisbon's response to the English mercantile community in the seventeenth century.
Next Article:From merchant capitalists to Corsairs: the response of the Muslim merchants of Malabar to the Portuguese commercial expansion (1498-1600).

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |