Toward the Reconstruction of Sacred Medieval Spatialities: Multiscale Analysis of the System of Eremitical Landscapes in Southern Portugal from Historical Records

Reconstruction of the landscape of late medieval southern Portugal is hampered by the scarceness of physical evidence after the subsequent centuries of evolution. The aim of this article is to understand how the spread of eremitical communities in this region interacted with the existing landscape, contributing to its transformation. A multiscale analysis is carried out through the examination of historical records containing scattered data about the landscape system. This methodology is applied to a specific case study, the Portuguese eremitical congregation of São Paulo da Serra de Ossa, in southern Portugal. Although fragmentary and dispersed, primary sources from between 1366 and 1578 are useful in reconstructing the components of this landscape. Initial results show the definition of a typical Mediterranean form of settlement: “agro-sylvo-pastoral.”


Introduction
If there has been difficulty in recovering knowledge of the previous state of the landscape due to its unceasing transformation over the last centuries and the scarcity of historical records and studies, such is even more the case with regard to the eremitical movements of the late Middle Ages in Portugal. Previous studies approach agricultural evolution in Portuguese territory (Cincinnato da Costa and de Castro 1900;Rau 1965;de Oliveira Marques 1978), and some address the southern regions specifically, focusing on the case of Alentejo's dehesa (Feio 1949;Parsons 1962;Joffre et al. 1988;A. Fonseca 2004;Fra Paleo 2010). 1 The only known monastic chronicles linked to the movement of hermits (de Santo António 1745; de São Caetano 1793) provide very little data about the structure of the landscape. Moreover, these chronicles are concerned with displaying a continuity between the eremitism of Serra de Ossa and that of primitive Christians, ending their narrative precisely in the middle of the 14th century, when the movement's first historical records appear. Many of their considerations are therefore apologetic and lacking an historical basis, even though, as Fontes (2012) highlights, the second of these sources offers some information useful for the reconstruction of the custodial history of some records. Importantly, in the 18th century Damásio lists documented records related to the mid-14th century. This means that, as Fontes shows, there has not been a great loss for this period in terms of documentation. Indeed, several documents are still available today. A few studies in the field of medieval history (Mattoso 1972;Beirante 1985Beirante , 2000Fontes 2005Fontes , 2012 have led to an increase in knowledge about the most representative movement in southern Portugal, known as the "men of poor life" (da pobre vida). Recent work has combined both historical and architectural approaches (Volzone and Fontes 2020). Despite this, a reconstruction of this medieval sacred spatiality has yet to be robustly researched. This article sets out to address this research gap by combining methodologies from history and landscape architecture. Our study of historical primary sources took an ecological perspective 2 -focusing on the way in which people and natural landscapes interact-in investigating the eremitical settlements: the organization of the properties and the administrative logic. We first examine the genesis of this eremitical movement and the spatial distribution of its communities in Portuguese territory. Then we delve deeper into the fragmentary and dispersed primary sources through the identification and analysis of 1,300 historical records. Of these, 300 were selected due to the paradigmatic data they contain about specific spatial components helpful in defining the structure of the eremitical landscape. After that, in order to systematize this information, the components are introduced into a database and grouped into "cultural" and "ecological categories." On the basis of these categories, four systems are identified: "circulation," "vegetation," "water," and "building." We discuss the importance of this analysis in revealing the transformation and construction of the landscape of southern Portugal by hermits; we also consider how this analysis represents a basis for future studies in the field of archaeology.

Historical Framework: An Eremitical Movement in the South of Portugal
This article considers the historical period from 1366 to 1578. A documentary source dated 1366, the oldest found to date (Fontes 2012), already attests to the existence of "men of poor life" in the Ossa Hill Chain. In 1578, the papacy integrated, with a large degree of autonomy in its government, the existing Congregation of the Hermits of Serra de Ossa into the religious order of Saint Paul the First Hermit. The analyzed time frame includes the historic convocation of the Council of Trent (1545)(1546)(1547)(1548)(1549)(1550)(1551)(1552)(1553)(1554)(1555)(1556)(1557)(1558)(1559)(1560)(1561)(1562)(1563), after which all the religious or mendicant orders went through a phase of reform and transformation. In a period marked by the Black Death-the first epidemic dates from 1348-with strong effects on Portuguese population, and by the famine associated with repeated years of poor harvest, the Crown was naturally motivated to develop a policy aimed at demographic and economic recovery and reorganization in order to attract population to new villages and municipalities (Beirante 1985). The plague and the population exodus resulted in large areas of abandoned and uncultivated land, so it is not surprising that municipalities supported hermits who would certainly contribute to the settlement and recovery of farmland. In addition, hermits were admired by different institutions and populations because of their exemplary way of living (Beirante 1985; Fontes 2012). The region chosen by "men of poor life" for their eremitical settlements presented particular natural conditions, along with a peculiar population, that clearly set them apart from the northern region of Portugal. Indeed, the southern region of Alentejo-whose name literally means "beyond the Tagus River"-is dominated by the peneplain. 3 The soil is not very rich in terms of hydric resources in comparison with the northern regions, with low rainfall and low-medium agricultural capacity (Rau 1965). There, human occupation has always been limited and concentrated, mostly because of the soil's low fertility, as well as historical and political factors (Feio 1949;Sousa 1988). The low population and the consequent shortage of labor favored the presence of large uncultivated areas distant from urban centers. These were used for pastoral transhumance and for cattle breeding, important sources of income through the sale of animals and related products: meat, milk, and fur (Morais 1998).
In the southern Portuguese territory, particularly in the diocese of Évora, it has been verified that there was only a thin regular settlement of monastic orders. Since the middle 13th century the mendicant orders, particularly the Franciscans and Dominicans, who enjoyed special papal support through several bulls (Vilar 1999), spread quickly in the area, with their convents near the main urban centers. It is in this context that the hermits of Serra de Ossa expanded and, later, became a congregation (Fig. 1).
Small communities of 3-7 hermits can be identified, 4 with exception of the community of Serra de Ossa and Vale de Infante, each with 20 hermits. These were governed by a mayor. This eremitism was closer to the Franciscan model, associated with pauperism, combining solitude and penance with manual work and forgoing pastoral activity. Of note is the isolated character of the places in which they settled, for which they are termed "hermitages." Indeed, they settled at the edge of the urban nucleus, where the environment could guarantee seclusion and self-subsistence. The communities did not accumulate many possessions, with exception of Serra de Ossa-the motherhouse-which had considerable assets. However, they tended to accumulate the properties needed for a consolidated area of possession, mostly around their settlement, either through donations, purchase, or by obtaining letters of sesmaria-granted to them by the municipalities on condition that the land was cultivated. These were important because the hermits 3 The region is bordered to the north by the natural boundary of the Tagus River and Ribatejo, to the south by the Algarve region, to the west by the Atlantic Ocean, and to the east by Spain. The peneplain, with a median altitude of ca. 300 m, represents a hinge between the lowest areas in the south of this region and the highest in the north. It is interrupted by few hill chains-São Mamede in the north reaches a height of 1025 m, while others are not very high: Ossa (698 m) in the northeast, Monfurado (423 m) in the west, and Portel (412 m) in the southeast. 4 A document (Congregation of Serra de Ossa [1573][1574][1575][1576][1577][1578]) sent in response to an inquiry ordered by the pope-drafted between 1573 and 1578-shows the number of religious men living in monasteries in that decade. could sell part of what they produced to people from the urban centers-wooden spoons, among other items.
In addition to the locus eremus or eremitério (hermitage), the Portuguese late medieval documentary sources use the terms "oratorio" (oratory), "cela" (cell), and "provença." The latter term is frequent in the older toponymy, and it sometimes survives in the more recent cartographies, making it possible to identify the settlements' locations. The term is not exclusive to Portugal, but appears in other geographies, such as Spain and Italy. Indeed, a broad movement, especially in southern Europe, proposed a return to the wilderness and poor life. De Viterbo (1865) includes two definitions for the term "provença": one refers to the spatial ideas of "province, region, diocese"; the second refers to "providence." It is interesting to note that the term "proveza" appears on the same page, also referring to "poverty." One record refers to the words "pobrencia" and "probencia" as closely linked to pobres or probes (poor men) (Rio Mourinho 1483a). The term can thus be understood as used with the meaning "place of the poor life." This can be compared with the first settlements of the Franciscan observance, which have been analyzed from a historical point of view by Vitor Teixeira (2010), José García Oro (2006), and Mario Sensi (1985. 5 Comparison can also be made to the settlements of the Castilian Hieronymites (dos Santos 1980; Ortiz 2021).
We have introduced the location of the eremitical settlements under consideration into a map of Portugal by cross-referencing data from historical records, previous studies (Fontes 2012), the analysis of cartographic sources, onsite visits, and interviews with local administrations and management entities ( Fig. 1). Here, as well as in other geographies, these settlements played an important role in periurban areas, working as dynamic elements of their surrounding areas through their spatial organization. Indeed, despite their ascetic practices, certain activities were indispensable for their self-sufficiency, e.g., raising cattle and beekeeping. For these reasons, they used the resources of the forests and cultivated lands around the hermitage, obtained through donations and sales that are documented with official charters. These documentary records are analyzed below, with the aim of reconstructing the landscape in which the "men of poor life" settled.

Toward a Reconstruction of the Eremitical Landscape from Historical Records
We approached the history of the construction of the landscape through the analysis of documentary sources, which, despite being fragmentary and dispersed, have been useful to identify specific spatial components that helped to define the eremitical landscape. In this case, analysis of primary sources is made from an ecological perspective, focusing on the interaction between hermits and the environment in which they settled. Although they bear little evidence when compared with the existing documents on other congregations or religious orders, charters for eremitical communities played an important role in the delineation of each eremitical settlement. The medieval landscape system that welcomed the "men of poor life" was reconstructed by elaborating the information contained in 1,130 charters, transcribed in previous research (Fontes 2012) and dating from 9 March 1366, including a charter for land donated by Gil Afonso to the Serra de Ossa hermits (Estremoz 1366), to 1578, when Gregory XIII (1578 issued the bull Creditum nobis for the approval of the Order of Saint Paul. These documents represent a corpus of evidence that makes it possible understand the eremitical landscape of the late Middle Ages. They were issued by kings, landowners, nobles, municipal authorities, and clerics, and they include such records as donations of properties, sales, and exchanges. Mainly conserved in the Évora Public Library, the National Archive of Torre do Tombo, the National Library of Portugal, and in district and municipal archives (e.g., those of Évora, Coimbra, Beja, and Portalegre), these sources include valuable data about the structure of the lands of the eremitical settlements, their boundaries and confines, with information about circulation, vegetation, and water systems. Upon analyzing the set of 1,130 charters, 300 were selected for the paradigmatic data they contain about the landscape of the eremitical settlements. They were introduced into a database, enabling categorization on the basis of the typology of the records. By cross-referencing the typological categories with the production dates of the documents, it has been verified that, generally, the first historical records regarding the communities belong to the donation charters. It is clear that, in a first phase, the expansion of the eremitical settlements was driven by land donations, mostly by local communities. Moreover, the hermits received important privileges, such as tax exemptions, from the Portuguese kings and benefited from their influence in gaining other privileges from the papacy. Regarding the foundation of the hermitages of Montemuro (Duarte 1415) and Penha Longa (João I 1390), among others, there was direct intervention by the monarchy in order to assure them the possession of the lands on which they sought to pursue their religious vocation. Beyond the Crown, the support of the municipalities is evident through the granting of lands in sesmaria, as is that of the local population through land donations. The nobility was not as involved, being more linked to other orders. However, the support of the Pereira and Bragança families is attested.
In a subsequent consolidation of the movement, communities started to implement strategies of property management that were also based on the acquisition of new lands, especially around the hermitage, in order to increase the land they could work and ensure their subsistence and isolation. Through an analysis in which the number of documents and the communities are related, it is possible to conclude that Rio Mourinho, Vale de Infante, Serra de Ossa, Alferrara, Valbom, and Aivado appear in a larger number of charters, mostly of the donation and sale type (Fig. 2). 6 It is not surprising that these communities, as well as others with a consistent number of charters, were not dissolved before the 16th century, instead becoming part of the Congregation of São Paulo da Serra de Ossa. 7 To account for this, a set of variables has been grouped into two categories based on the division elaborated by Magalhães (1996): "cultural" and "ecological" components.  Rolando Volzone, 2020.) This allows the definition and delimitation of the landscape's structure, and it contributes to further knowledge of the history of humanization and construction there. The ecological components consist of the hydric system, spontaneous vegetation, orography, and geology. Cultural components represent the way cultures use ecological resources for daily life: equipment, land use, and hydraulic systems. Due to the kinds of data available and through the interaction between the components, this study identified four systems: circulation, vegetation, water, and building (Fig. 3). The spatiality is further explored through a multiscale analysis process: the study proceeds from territorial scale to hermitage scale. Therefore, this study begins by analyzing the roads and pathways-the first way to access the land-then moves to explore the lands themselves and their use, the enclosure and its interior, and, finally, the buildings.

Circulation System
Hermits intended to lead a secluded life. Distance was important, but also relative, as the microcosm they constructed was not a space of complete isolation. The fenced perimeter that often surrounded lands and hermitages welcomed the outside world, since the hermits' tradition required them to provide hospitality to those who were looking for it. However, the opposite happened too: hermits visited and stayed in cities when they needed to celebrate a juridical act, request a copy of a document, or simply to deliver surplus fruits from their lands or the wooden spoons they made to women who would sell them (Fontes 2012). For these reasons, a strict relationship between the hermitage and the system of circulation was necessary. Based on historical studies of the Alentejo region, it is known that some Roman routes were used during the medieval period, and others were In some cases it is evident that the existence of a road could determine the selection of a place. In other cases locations led to the creation of new roads. This defines a system enriched with other hierarchically organized connections (Volzone and Fontes 2020). By overlapping the ancient network of Roman roads, adapted from previous studies, and the location of the hermitages, even at a macro scale, it is evident that the eremitical settlements lay in areas relatively close to the roads (Fig. 4). The main Fig. 4 The relationship between the network of ancient Roman roads in Portuguese territory (data adapted from Soutinho [2020]) and the locations of the hermitages. (Figure by Rolando Volzone, 2020.) road network generally ensured connection with the most important nearby settlements and with the rest of the territory. A secondary network of paths-which were sometimes narrow, allowing the passage of just one person or two-connected parts of the same settlement, such as orchards and vineyards (J. Fonseca 1998). Indeed, some lands needed to be worked on a daily basis, and a network of rural paths was indispensable to ensure capillary circulation through them. These two levels of road network outline the landscape, and, together with other components, serve to delimit the land possessions, whether larger or smaller.
Analysis of the charters helps in understanding the various diverse elements useful as spatial delimiters and in reconstituting the Alentejo regional landscape of the late Middle Ages. Regardless of their dimensions-a type of information that is rarely included in the records-the location or the type of use of hermitowned lands needed to be delimited in order to facilitate their management. Previous studies (Gonçalves 2013) argued for the use, at first, of less-defined elements materialized in landforms, such as ridges, heads, waterlines, and valleys. These delimitations have subsequently acquired a greater accuracy, thanks to continuous elements-e.g., ditches-or punctual elements, isolated or in sequence-e.g., markers. The documentation designates an abundant set of physical elements for the lands' delimitation into spatial units. Sometimes these were elements created specifically for this purpose: boundary markers, as stemming from the Roman tradition, were usually positioned in series, along paths, or at the limits between two lands belonging to different owners, sometimes underlining other types of property markers, such as ditches. 8 Ecological components were also able to serve this function naturally, without human intervention. Brooks (Vale da Infante 1437) 9 and streams (Rio Mourinho 1483b) 10 were suitable as continuous property delimiters (Gonçalves 2013).

Vegetation System
The vegetation system is the result of the interaction between ecological components that are natural to a specific area and cultural components, introduced through human action. Both are morphological components that define the identity of a landscape. In the late Middle Ages in the Évora area, for instance, land use was spatially distributed in a concentric form around the urban centers. Cultivation that needed daily attention occupied areas closer to the city, within walking distance. Agricultural systems followed a typically Roman organization of the landscape, split into productive and complementary groups: urbs, hortus, ager, saltus, and silva (Mazoyer and Roudart 2001), corresponding respectively to the urban area, kitchen garden, agricultural field, pasture paths, and forest.
When applied to Portuguese territory, four concentric circles are identified (Barata and Leite 1996). 11 The first of these-about 1-2 km from the center or from its walls-was that of the ferragial, an intensive livestock feed-production site where barley and rye were cultivated. Other forms of agricultural production were common as well, such as kitchen gardens in a rotating system, in order to ensure food sustainability for the city. This circle was located nearest to the city because of daily needs. The second circle was located from 2 to 5 km away from the center and was generally occupied by vineyards, which, unlike the kitchen gardens, did not need daily care over the whole year. It was also cultivated with other arable crops, depending on the ecological conditions that were more apt for certain products. In the Alentejo region, vineyards were complemented mostly with olive trees or figs. Even from the landscape and morphological point of view, they constituted a boundary of the peri-urban space (Serra 2006). Indeed, Langhans (1970) also notes that the cultivated areas were those in the immediate surroundings of the villages, and the rest was endless scrub and desert. Farms were located in the third circle. There, crop fields and grazing lands coexisted. These areas did not need daily treatment, so they were located 15 km away from the center. The last circle was for pasture. Évora had important indigenous cattle production and was also a place that transhumant livestock passed through on its way from Serra da Estrela to the southern fields and in their return to the highlands in the summer (Morais 1998). Finally, there are two more systems to highlight: harvest lands 12 and woods. 13 The former could be associated with the ferragial, vineyards or parcels of the farm; the latter had an important role in the population's daily life, as place of feed, fodder, and raw materials. Monoculture fields with an exclusive system were rare-indeed, associations were common: kitchen gardens with harvest lands, and vineyards with olive trees, cork trees, or orchards, among others (Barata and Leite 1996). Starting from the distinction between ecological and cultural components, we have grouped all the identified landscape data regarding the vegetation system into the seven categories previously analyzed. In order to correlate these components and their presence in 17 eremitical communities, 14 a radar-type chart was made. This chart supports different interpretations. Firstly, it makes it possible to understand the presence of each component in the settlements, enabling a specific comparison. Secondly, it makes it possible to verify the global richness of each community in terms of its agricultural products on the basis of the documentary references, taking all the components into consideration simultaneously. Thirdly, it highlights the association between systems of grinding equipment in specific settlements and components of land use (Fig. 5).

Water System
The wide availability of water-part of the hydric system-enabled hermits to carry on their secluded life with the necessary autonomy from the closest settlements. The hermits needed a hydraulic system to collect, transport, store, and distribute water for personal use (washing and cooking), as well as for the irrigation of crops and for cattle (Mascarenhas and Jorge 1996). In addition to nourishing the earth, water played a religiously symbolic role in the hermits' settlement (Volzone 2020). Consequently, it is not surprising that hermits settled near springs or important watercourses. The names of rivers are revealed in several documents, and it is interesting to verify how many of these hydronyms are still present in contemporary cartographies. In several records the community's name itself includes the hydronym: Rio Mourinho, Rio Torto, Água das Infantes, and São Brissos de Xarrama.
The energetic potential of water was also used as a driving force for the operation of grinding systems: mills, watermills, and trip-hammer mills. Mills were the most-used technologies for raw-material processing and transformation. In Portugal, as highlighted by previous studies (de Oliveira et al. 1983), there existed two types of grinding devices moved by hydraulic power: the moinho-mill, generally with a horizontal waterwheel-and the azenha-also known as a watermill, using a vertical waterwheel. 15 These 12 The Alferrara community charters refer several times to harvest lands: "herdade de pam que eu ey onde chamam os barris termo da vijlla de pallmella" (a harvest land with cereal that I have, called "os Barris," on the outskirts of Palmela) (Alferrara 1462, 1520); "casal de pam que Elles avijam a beira da mata do barrill" (a harvest property with cereal that they had next to the wood of Barril) (Alferrara 1475); and "courela de herdade de pam" (a small piece of harvest land with cereal) (Alferrara 1476). Sometimes they are associated with olive groves, but most of the time they relate to the presence of woods. Around 75% of documents with references to harvest lands belong to sale charters. In one of these, harvest lands are associated in the same land with pine forests, olive groves, and a quarry with an adjoining limekiln (Alferrara 1494). 13 References are mostly found in relation to the Mendoliva and Alferrara communities in the Setúbal Peninsula: "mata de el rei" (wood of the king) (Alferrara [1465]) and "mata do Barril" (Wood of Barril, in relation to the hill with the same name) (Alferrara 1475, 1476). In the first identified charters, the wood was located immediately next to the hermits' possessions, delimiting the hermitage. Since 1490, sale charters show part of this wood being purchased, demonstrating the enrichment of the communities and the accumulation of land around the hermitage (Alferrara 1494).
14 These 17 case studies were chosen for deep analysis due to two different factors: firstly, these eremitical communities have developed into monastic ones; on the other hand, these tend to present a more consistent number of charters, being paradigmatic in terms of landscape data. 15 According to de Oliveira et al. (1983), windmills would have been rare in Portuguese territory between the 14th and 16th centuries. needed to be located in places with abundant watercourses 16 in order to work. They were distributed all over Portuguese territory, depending on the local ecological component and the proximity to urban centers or to the circulation system. As with the analysis of the vegetation-system components, we have created a radar chart to display the relationship between the components of the water system and their presence in records regarding a specific settlement (Fig. 6). This makes it possible to interpret the data of a specific settlement, while at the same time enabling comparison with the others. By analyzing these data, it is evident how communities around the Ossa, Arrábida, and Monfurado hill chains were richer in terms of hydric and hydraulic components. Through this analysis, the majority of the records in the database regard the Serra de Ossa, Vale da Infante, Valbom, Alferrara, and Rio Mourinho communities. This shows a correspondence with the chart for the vegetation system, underlining the link between water resources and soil fertility. The communities of Serra de Ossa, Vale da Infante, and Rio Mourinho show multiple references to água (water). 17 Since most of the references regard running or flowing water, it is evident that they occur mainly in mountainous or valley topographies. In some cases-above all in Rio Mourinho-the presence of records regarding grinding equipment is strictly linked to the prevalence of references about the water system. In other cases, this correspondence does not exist, as evident in Junqueira, Serpa, and Rio Torto. As seen before, the geomorphology influenced eremitical settlements through the interaction among orography, vegetation, and water-basic components of any natural place. Understanding these factors, the hermits respected and accommodated the genius loci (Norberg-Schulz 1980). Each settlement needed a natural framework with specific features. These features met practical needs, such as the population's relative remoteness and the presence of abundant drinking water, which also supported the fertility of the soil. The spaces chosen were characterized by the presence of (1) hills, (2) valleys, and (3) important waterlines (Fig. 7). These elements were included in charters several times. The Alentejo region is mostly flat, considered to be a peneplain interrupted by chains of low-altitude mountains or hills. Detachment was the most important element; in those centuries, hills as well as valleys were not very busy places. For hermits the desire for seclusion implied distance from community life, which provided the tranquility necessary for this life of meditation, prayer, asceticism, and silence in contrast to life in the outside world.

Building System
Concerning the architectural elements forming the hermitages, records are extremely scarce. The  Rolando Volzone, 2020.) systematization, critical analysis, and interpretation of the few clues included in the historical records allow us to outline some conclusions aimed at identifying the physical structures that housed the hermits or fulfilled certain functions. As seen before, grinding equipment was identified in several settlements. In this regard it is interesting to note that the hermits were involved in the repairing and maintenance of these structures, such as watermills or trip-hammer mills, which were built close to the watercourses. This emerges, for example, in records referring to the hermit Gonçalo da Silveira and his companions. 18 It has therefore been supposed that hermits were involved in the construction of these and, perhaps, other elements. Although local materials such as stone were used for some equipment, these medieval constructions were mostly made of perishable materials, especially wood, and so were quite fragile and rudimentary. For this reason, several references to ruined buildings, sometimes completely destroyed, are identified. The pardieiro or paredeiro is a physical structure that emerges in some records. The term refers to a house in ruins, deserted or uninhabited, indicating the presence of a previous dwelling that was abandoned afterwards. In Cela Nova this term is associated with harvest lands-terra com pardieiro (Rosa 1516a, 1516b). In this sense it corroborates the abandonment of agricultural land, when associated, for example, with dead vineyards or lands previously cultivated and then overgrown, in Valbom (Valbom 1416). It justifies the interest shown by municipalities and kings in promoting the cultivation and settlement of these lands, which were also donated for this reason. But these structures are rarely mentioned as being within the enclosures of the hermitages, instead being mostly associated with their areas for cultivation. Through the analyzed records, the existence of portable altars is verified. The presence of these structures has been documented since 1402 through the bull Sincere devotionis affectus of Pope Boniface IX. Hermits obtained the license to own a portable altar and to erect a belfry for Masses or celebrations of other services by a suitable priest of their choosing (Boniface IX 1402). Consequently, hermits did not need a built chapel, but could adopt this type of structure to celebrate anywhere. Fixed altars were gradually introduced with the consecration of worship places inside the hermitages, as affirmed by Fontes (2012:201) (1483) about a conflict between "men of poor life" and Hieronymites includes a description of the incident from both points of view. 19 Despite the differences in their stories, common spatial elements are identified in the hermitage Serra de Ossa. Firstly, the record mentions the enclosure (cerquoyto) with a possible entrance. Secondly, the presence of an oratory can be verified. Thirdly, there were houses in which the hermits took refuge, closing the doors. This proves the existence of a built structure, not just a cave typical of the eremitical traditions. Similarly, in a charter dated 1483 regarding the resolution of the conflict, once again, between the "men of poor life" of Rio Mourinho and the Hyeronimites, a description of assets contains data on the architectural system of this community (Rio Mourinho 1483b). This consisted of the mud wall, some grinding equipment, an oratory, and also some houses; even more interestingly, the presence of six beds, including their linen, is also noted. It is likely that the same typologies were associated 18 "E que ao tempo que elle o dito mojnho ouuera, Era muito delapidado, e que ele com os outros pobres seus parceiros que no dito logo com ele estauam trabalharam no dicto muinho e ho lançarom a moer" (And by the time he had the mill, it was badly damaged; and he, along with the other poor, his partners, who lived with him in that place, worked to rebuild the mill and managed to get it working again) (Afonso V 1447). In addition, a sesmaria charter allows the construction of a watermill close the Xarrama Stream (São Brissos 1409). 19 With regard to the hermits: "[S]se meterom dentro em ssuas casas e oratório e çarrarom de dentro ssuas portas e as trancarom muy bem, E os dictos dous frades e vigayro e os homens de cauallo sse forom as portas pera as abrirem E botarom duas fora do couçe e per força entrarom dentro contra a vontade dos outros yrmitães que na dicta casa Estauam" ([T]hey went inside their houses and oratory and locked their doors from within and locked them very well. And the two friars and vicars and the men on horseback went to the doors and tried to open them. And two doors burst open with the kick of the horses, and by strength they managed to enter inside against the wishes of the other hermits who lived in the community). The Hieronymites' version differs: "[S]sem outra força foram dentro ao oratório e casas da dicta provençia estando as portas abertas por onde entraram paçifiquos e manssos ssem outro escandallo" ([W]ithout any force, they went into the oratory and houses of the province, the doors being open, through which they entered peaceful and meek, without any scandal") (João II 1483). Underlying the difference between the monastic and the eremitical typology even more, in the same document they declare: "Elles nom conssentyam que ssua casa e oratoryo fosse tornado em moesteyro" (They do not consent that their house and oratory are turned into a monastery). with the other eremitical settlements. Indeed, already since 1383, a charter concerning the Rio Torto community 20 attests to the presence of closed houses (Rio Torto 1383). It is therefore possible to deduce that the model adopted foresaw the existence of cells, or houses, and not a common dormitory. This kind of shelter populated the landscape even in the engraving contained in Chronica dos Eremitas da Serra de Ossa (de Santo António 1745), which represents the Serra de Ossa monastery on a slope of the Ossa Hill Chain. This could indicate the existence of small oratories inside the hermitages, even after the institutionalization of the eremitical movement. As seen previously, due to the values of the eremitical tradition, "men of poor life" welcomed old and sick people and pilgrims, but also men interested in buying products, such as wooden spoons or fruit. It is evident that structures for hospitality existed, as proven by a testimony given by Brother Paulo de Portalegre in Novo Memorial do Estado Apostólico-the first chronicle of the Secular Canons of St. John the Evangelist, dated 1468. 21 This document contains the description of the welcome received by Brother Paulo de Portalegre in the Serra de Ossa hermitage, and it mentions the presence of a structure for keeping horses and of houses inside and outside the enclosure. On the one hand, this implies a differentiation of spaces; on the other, it implies that not everyone could enter into the hermitage's most intimate space. Information about the architectural system is mostly fragmentary and sparse. This hampers a more complete reconstruction. However, it is very probable that the identified built components were replicated in the other hermitages.
It is possible that gradual spatial organization, still far from following a defined model, was their basis. In the Ossa Hill Chain, hermits could firstly have taken advantage of the natural resources for shelter. They most likely settled into caves, such as those located in Vale da Infante (Fig. 8), or made use of other structures already existing on the lands (Boaventura et al. 2014). 22 It is important also to highlight that the papal bull Cum vobis visitationem of Pope Gregory XI gives precious data about the communities existing in the entire Portuguese territory. In fact, a visit was ordered to inquire about the life, customs, and orthodoxy of these hermits. On the 2 February 1378, Vasco Domingues issued a statement focused on these hermits, who lived in "mountains and desert places of the Évora Diocese and Alentejo County" in groups of four. 23 However, the lack of historical records that indicate the exact location of this kind of occupation, as well as of in-depth archaeological studies, do not allow us to confirm this theory yet, and the evolution and spread of the movement ended up not pursuing such occupation.

Discussion and Conclusions
Landscape construction arises from the interaction between humans and the environment. The eremitical landscape of "men of poor life" and the construction of their symbolic spatiality is no exception. The locus eremus is defined in terms of practical and spiritual aspects, and those related to the spiritual dimension are strictly linked and guaranteed only if the material needs are satisfied. Our analysis of the paradigmatic data included in the medieval charters has distinguished cultural and ecological components. Our study of cultural elements has verified that their presence at a specific site was due to the natural substrate, soil, and geology. Indeed, the density of a specific cultural component is a function of the presence of certain ecological components.

"[D]erom aos pobres e a el e seus Jrmãos que ujuem em na dicta proue ujda o dicto logar em que ora estam com suas
Casas çaradas com suas aruores em que fazem serujço a deus" ([T]hey gave the poor and to him and his brothers who live in poor life the place where they are settled with its houses, enclosed, and its fruit trees, where they serve God") (Rio Torto 1383). 21 See the critical edition, de Portalegre et al. (2007). 22 The archaeologist Rui Mataloto, who carried out the excavations in the Ossa Hill Chain, confirms that the dolmen of Candeeira, with a peculiar hole in its structure, was probably used by medieval hermits. This can be confirmed by the location of that dolmen and by the material evidence that emerged during an archaeological campaign in 2013. Indeed, the dolmen was located close to a natural pathway-still marked in the Portuguese cartography of the 1960s and 1970s-connecting the various routes that made it possible to cross the Ossa Hill Chain in a north-south orientation. However, an estate named "Candeeira" appears in written records only in the 15th century as one of the hermits' properties worked by caseiros (lay peasants) with no reference to a previous eremitical settlement (Afonso V 1472). Exemplifying this relation is the presence of limekilns and glass furnaces in correspondence with limestone or sandy areas, respectively. Indeed, the hermits showed great empathy to the place and deeply respected its genius. Charters contain references to cultivated land inside and outside enclosures that were farmed by hermits themselves, along with the presence of cattle and beehives, displaying an agrarian distribution typical of periurban areas. Similar to other monastic experiences, the more fragile cultures (culturas mimosas, in Portuguese), such as orchards and vegetable gardens, created a first ring around the hermitage. These were often inside the first enclosure because they required more maintenance and surveillance. This also justifies the installation of wells and reservoirs in such areas. Vineyards and olive groves formed the second ring. It should be noted that these settlements-despite not yet having a monastic structure-had an administrative logic that was not only spiritual, but also revealed humanization of the land. Hermits were extremely clever in the way they organized the properties obtained through donations of lay benefactors and municipal authorities or through purchases. Some communities accumulated considerable possessions, increasing their acreage around the hermitages by creating a compact area around the enclosure. This allowed them to safeguard their seclusion and subsistence while enlarging their cultivated areas. They sometimes owned areas that may even have been geographically isolated, but that ensured their survival. In such cases, the lands farthest from the hermitage were assigned to the care of a tenant. Their settlement in scarcely populated areas contributed to the human occupation of the surrounding rural municipalities and the enlargement of their cultivated areas. On the one hand, they used local resources (e.g., water, wood, and stone); on the other hand, they transformed the surrounding landscape with the support of those municipalities that granted lands in allotment, knowing that they would ensure their cultivation and use. Moreover, through their agroforestry-pastoral system they developed their own subsistence economy, but maintained relations with the outside world: they sometimes needed to go to the city or to welcome those seeking their hospitality into the hermitages. These foundations brought something new to the spatial framework of the rural monastic experience of the late Middle Ages. Indeed, by virtue of their spiritual vocation, they changed the space while, at the same time, adapting themselves to the space and the environment. Additionally, by cultivating abandoned areas away from the urban centers-areas where they could take advantage of existing resources, such as water, soil, and vegetation-they introduced a new typology of settlement into the south of Portugal, one different from that of the urban mendicants found in the north.
Another interesting conclusion is that the presence of mountains and valleys tended to play a role in site selection in view of the territory's ecological richness, the isolation these environmental features permit, and their symbolism. Unfortunately, in the oldest documents, the few indications about built elements are very dispersed. Although no references to a conformed space, following either the monastic or the conventual model, emerge until the 16th century, a gradual organization in terms of spatiality is verified closer to when this movement was institutionalized. This organization is evident through references to the delimitation made by fences, as well as to oratories, cells, and spaces of reception. No further elements exist that could provide a more complete image of the hermitage.
The empirical results and contributions of this study should be considered in the light of two limitations: (1) the scarcity of documentary sources and (2) the profound subsequent transformation of the settlements in terms of architecture and landscape. Due to their intrinsic nature, based on isolation, silence, and introspection, eremitical communities present a scarcity of documentary sources when compared with other religious experiences-even more so before 1378--as previous studies attest ( Fontes 2012). Indeed, eremitical communities followed no kinds of rules or norms until an organized congregation was established, and even then their statutes or rules do not contain detailed or prescriptive information in terms of spatial organization. Moreover, iconographic sources and technical drawings are rarely found, thus hampering the diachronic reading of certain buildings. Secondly, the conditions of conservation and accessibility are precarious for some monasteries that lost their religious designations. The great transformations these spaces have undergone over the centuries and their state of abandonment or decay often hinder a deeper analysis. Indeed, it is important to take into consideration that some buildings have partially or totally disappeared, sometimes leaving only few traces. Landscape analysis on the basis of contemporary findings is similarly difficult, as various transformations have compromised the edaphoclimatic conditions, e.g., eucalyptus forestation in the Ossa Hill Chain.
This research adds new information to the organization of the eremitical landscape structure. These kinds of settlements are commonly imagined as devoid of real constructions. However, the analyzed data show that the landscape was really altered through the introduction of orchards, kitchen gardens, vineyards, the hydraulic system, and so on. Indeed, the eremitical landscape was not reduced to pure experience and adaptation to the landscape, but was constructed. The research on this topic is still open, and this work, by establishing a dialogue with existing studies, can serve as a starting point for future research. Future archaeological studies should be undertaken to complement the data exhibited in this research with material evidence from onsite works. It is important to investigate the sites of the medieval settlements in order to understand how they are spatially organized and what structures exist (if they still do): enclosures, interior and exterior spaces, dormitories, cells, common spaces, etc. Specific cases are identified in the areas around Portel (Água da Infante hermitage), Serpa (Serpa hermitage), and Portalegre (Vale de Flores hermitage). Secondly, fieldworks in cave eremitical settlements should be analyzed further. In Vale da Infante, for example, three caves have been identified, suggesting that other caves must have punctuated the hill chains described in the historical documentation: Ossa, Arrabida, Monfurado, and Portel. Identification, spatial analysis, and survey of these would make it possible to understand their relationship with the network they formed and with the whole territory, and to overcome the scarcity of information provided by the available documentation. Thirdly, the current location of the monasteries could, in some cases, reveal medieval structures, taking into account enclosures, mills, watermills, the network of pathways, and the hydric system, among others. Structures have been identified in the Ossa Hill Chain (a watermill close to the hermitage of Santa Maria) and in Palmela (a limekiln in the enclosure of the monastery Alferrara), but archaeological studies are needed.
Moreover, it would be interesting to analyze these communities' landholdings in Portuguese territory, transposing the data into a georeferenced system aimed at understanding the real extension and the diachronic expansion of their assets.

Conflict of interest
The authors declare that there is no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.