Introduction

Landscape and territory are as closely linked as the territory and its cartographic representation. As a support for the representation of the territory, geographic information can be a vehicle for identifying common or discriminating elements in the landscape study. From this perspective, cartography and its toponymy share variables with landscape management tools, being transversal tools in the multi-layered territory analysis (Nowosad and Stepinski, 2021).

Place names determine geographical units in space, synthesize the description of the same places they nominate, and treasure the perception of the territory through time. This description of toponymy is fully aligned with the concept of landscape. The Council of Europe signed the European Landscape Convention (ELC)Footnote 1 in 2000, setting a milestone in the recognition of the value and importance of landscape. This treaty provides a framework of protection and orients the concept of landscape towards the land use planning field, which has greatly impacted the scientific literature (Pătru-Stupariu and Nita, 2022). In said convention, the landscape is understood as ”an area, as perceived by people, whose character is the result of the action and interaction of natural and/or human factors.” This definition has a bearing on two concepts intrinsically linked to place names: human action and perception. Hence, both toponymy and landscape become guardians of the cultural legacy (Altaba et al., 2022).

Furthermore, several disciplines are involved in the study of geographical names, such as anthropology, sociolinguistics, geography, or history, among others, for which toponymy is an auxiliary tool. Moreover, reciprocally, toponymy is nourished by all these sciences, which it also serves. Therefore, toponymy is a bridging discipline between all this knowledge, and its study requires an interdisciplinary approach. One of the values of toponymy is its memory (Çetin, 2018), that is, the value of containing geographically located chronological information, as long as the origin, meaning, and reason for the timely toponym are unraveled. This is also understood as etiology, which provides knowledge of synthesis (Arroyo Illera, 2009). The confluence of different languages in the same territory, multilingualism, generates a particular interest in the study of place names. This happens in several regions of Spain, including the Autonomous Community of the Basque Country, or Euskadi, an administrative region in the north of Spain (bordering France).

Pathways, those communication routes that allowed transportation and connections between different places of human activity development, have been pivotal elements of the territory (Porcal Gonzalo, 2011). Therefore, the analysis of elements related to the traditional transportation network provides a basis for its interpretation. According to Porcal Gonzalo, the traditional pathway can be understood as a dynamic type of cultural landscape, both due to its spatial scope and its representation of territorial exchange processes, illustrating human adaptability to the geographical environment.

From the perspective of the genesis of cartography, roads and pathways have been the structure that enables the cartographer to establish contact with the terrain under survey as a means of access to the territory. In this way, they condition the perception of the space on which the work of abstraction and synthesis has been performed in the compilation of the maps using classical techniques. This process is evident in articulating the topographic works of the first edition of the National Topographic Map of Spain (MTN), produced by the current Instituto Geográfico Nacional (IGN) (Instituto Geográfico Nacional, 2024a), the Spanish national mapping agency.

Regarding the urban environment, roads give support to its framework, modeling the morphological structure of the population settlements, together with other elements such as river courses or orographic features. Examples of these constrained structures can be the layout of buildings linearly on both sides of a road, the concentration of small nuclei at their crossroads, or the design of radial planes in different directions (Agboola et al., 2018). In addition to structuring the territory, pathway infrastructures also determine the landscape. Thus, once again, cartography and Geographic Information Systems (GIS) are indisputable supports for assessing these relationships.

Additionally, UNESCO recognizes several roads and pathways as world heritageFootnote 2. Based on the ELC, the European Council drives the program of cultural itineraries (Council of Europe, 1987), among which various traditional routes can be found, such as the Via Regia, the Via Francigena, or the Camino de Santiago (Way of St. James). This program promotes the dissemination and enhancement of roads, as well as the reuse of disused transport infrastructures. Thus, it fosters an approach to rural environments that contribute to their socio-economic development, simultaneously meeting the increasing demand for eco-recreational countryside use. In Spain, the Natural Paths Program pursues similar objectives, focusing on natural landscapes (Orellana Moraleda, 2017). For example, Somoza Medina et al. (2022) analyze the successful case of the Camino de Santiago.

Delving further into the landscape-pathways pairing, it is worth emphasizing that the regional regulations (Departamento de Medio Ambiente y Política Territorial, 2014) outline the objective of ”enhancing the value of traditional cultural pathways as exceptional points for accessibility and enjoyment of the landscape.” Within these guidelines, pathways also find a place at various junctures. Firstly, in defining the content of the community’s landscape catalogs, the spatial location of main routes and viewpoints from which the landscape is appreciated emerges. Additionally, there is a requirement for a ”Landscape Integration Study” in construction projects that might significantly impact the landscape, particularly in the development of transportation or port infrastructures. Based on everything discussed so far, the authors believe that the significance of the relationship between the landscape and other elements within it, such as pathways, has not been addressed to date in Spain, especially in a bilingual environment.

This research aims to explore the relationship between place names and landscape units in the Basque Country. To this end, it is based on the idea that the perception of the occupants of a territory and their activity on it are the genesis and the driving force in the dynamics of both toponymy and landscape. Specifically, and given that traditional communication routes are structuring elements of the territory and the landscape itself, the work focuses on certain features that characterize the toponymy of pathways. Since the spatial component is the core of the analysis, it is carried out using GIS tools. Firstly, a toponymic corpus is compiled from the geographical information of the administrative entities related to the pathways in this territory. It must be noted that the toponyms related to communication routes are also known as odonyms. Finally, the analysis of these names, superimposed on the existing cartography of the landscape in the area, seeks to verify the correspondence between the differentiating characteristics of the landscape units and these names. For this purpose, the features taken into account are the distribution, geographic, and semantic referencing of the names.

Therefore, the main innovation of this research lies in systematically intertwining disciplines that work with the perception of the territory by its inhabitants through the analysis of elements that articulate or structure it, such as the pathways throughout the Basque Country territory.

Related work

This section presents a series of works related to the different disciplines that combine the study: landscape, toponymy, and traditional communication routes. Thus, it will show their close interrelation and the treatment they have received in the literature.

First of all, it is worth mentioning that Spain ratified the ELC on November 26, 2007 (the Spanish Official State Gazette, Boletín Oficial del Estado-BOE of February 2, 2008), which came into force on March 1, 2008. The effect of this implementation on landscape protection and management is reflected in Civitarese Matteucci and Franco Cartei (2022). The cultural landscape finds a specification in the rural landscape, characterized by natural components, diversity, and visual qualities (Picuno et al., 2019). The management and revitalization of linked rural areas are gaining more and more relevance, as evidenced by the European Commission’s initiative ”Long-term vision for rural areas of the EU up to 2040” (European Commission, 2011). Among its contributions is the establishment of 10 common objectives and the articulation of an action plan (Munroe et al., 2019).

The characterization of landscapes and their classification on the territory is another related and necessary subject in the areas of landscape management and planning contemplated in the ELC as a specific measure. Geographic information and cartography constitute support for its study through GIS (Tort-Donada and López-Leiva, 2023). In any case, the large number of variables and the complex relationships between them mean that landscape should be addressed as a broad and multidisciplinary field (Altaba et al., 2022). Usually, these studies center on physical elements and their visual component. In this aspect, analyzing place names provides the cultural component with which they are impregnated (Atik et al., 2022).

Furthermore, literature gathers some works related to odonyms. For instance, Bellosillo (1988) focuses their research on the study of ”cañadas reales” (Royal Drove Road, transhumance route) through their topo-nymy, placing special emphasis on the phenomenon of transhumance and the impact of wolves within the area with merino sheep in Castile region, Spain. Another notable work is the compilation of terminology related to pathways in Spanish language and dialects, which is based on the generic part of the names of these geographical entities (Alcázar González and Azcárate Luxán, 2000).

If the perspective is broadened to the concept of landscape, it is also possible to find examples related to the cultural aspects of pathways. Spampinato et al. highlight the possibility of analyzing the richness and diversity of land cover in the past by studying phytonymy or place names related to plants (Spampinato et al., 2022). Hearn et al. (2024) interpret the change in certain rural landscapes through the toponymy related to vegetation, delving into the imprint that nomadic culture leaves on both the landscape and its toponymy. Gordova (2022) links landscape dynamics with the chronology of place names, relating their linguistic origin to the settlements of different peoples in the territory throughout history. The stratification of toponymy allows a similar approach, as seen in Martínez-Areta (2023). Nowosad and Stepinski (2021) connect the characteristics defining the classification system of landscape patterns with natural variables and place names. After registering place names, it is concluded that environmental aspects are mostly divided into factors of natural or human geography. Focused on the latter, recent studies use innovative methodologies with toponymy as the axis for the interdisciplinary study of architectural heritage, as in the case of Herzen et al. (2023), dedicated to analyzing Russian religious heritage. Membrado Tena and Fansa (2023) apply a semantic analysis of urban toponymy in Valencia (Spain), examining the persistence of rural landscapes within it. Another example of applying technological advancements, centering on artificial intelligence techniques, is the semantics of place names, which have also been used for geocoding place names themselves (Fize et al., 2021).

In Euskadi, different territorial planning figures have been developed in which the protection of cultural heritage from a territorial perspective has been covered in one way or another. In the region, an integral approach has been adopted. Thus, beyond the protection of a specific cultural heritage element, action is taken on its location and its surroundings, thus protecting the landscape context (Davila-Cabanillas, 2021). (Hersperger et al., 2020) stress the importance and applicability of landscape knowledge applied to territorial planning. This issue connects with the relationship of the 17 Sustainable Development Goals with territorial planning from a landscape approach in order to provide the rural environment with greater resilience in the face of transformation (Riva, 2020).

From another perspective, both pathways and infrastructures also condition the landscape (Lisiak et al., 2018). In this way, the temporal development of the settlement of the territory is reconciled with the geography and justifies the existence and evolution of these paths (Rosete Vergés, 2020). Thus, attending to modern dynamics, new residential and infrastructure developments generate discontinuities and provide an irregular plot in the forest cover of the territory (Munroe et al., 2019). Increasingly, traditional paths also seize value as part of local heritage and a result of history and culture. This perspective relates to the concept of cultural itinerary (Álvarez et al., 2022). Turner et al. (2023) show an interest in historical inventories of roads and a way to obtain them employing historical mapping. In addition, they have aroused great interest as a tourist and leisure resource, especially in mountain areas, enhancing their scenic content. Other axes of great value for their resources are the spaces linked to the riverbanks and riverbeds. Their patrimonial, economic, and cultural wealth is an essential asset for developing the rural areas where they are located (Cebrián Abellán and García Martínez, 2016).

Fig. 1
figure 1

Study area. Euskadi / Basque Country Autonomous Community. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

Study area

The geographical location of this research is the Autonomous Community of the Basque Country or Euskadi, in the northern part of the Iberian Peninsula, bordering France, and adjacent to the Cantabrian Sea. It spans an area of 7234.83 km2 and has a population of 2,186,517 inhabitants (inh) (Eustat Agentzia, 2023), resulting in an average density of 302.22 inh/km2. It is composed of three provinces, whose capitals are indicated in parentheses: Gipuzkoa (Donostia/San Sebastián), Bizkaia (Bilbao), and Álava (Vitoria-Gasteiz, which is the capital of the autonomous community), as shown in Fig. 1. For historical reasons, these provinces enjoy a special regime according to their own laws (fueros), thus being recognized as Historical Territories.

Fig. 2
figure 2

Relief map of the study area. Own elaboration from geoEuskadi (2023) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Concerning topography and climate, the former is mainly mountainous, with a series of elevated orographic structures oriented from north to south, which organize the territory (Fig. 2). To the north are the coastal mountains, forming an alignment cut by the fluvial systems of the Cantabrian slope. To the south are the interior mountain ranges, which form part of the Atlantic-Mediterranean divide. Next, the middle pre-Pyrenean depression is developed, which forms the Álava plain (La Llanada Alavesa). Finally, the outer mountain ranges of the south are the last part of the territory. This orographic formation, along with the proximity to the sea and the region’s latitude, shape the climate of the region. According to the Köppen-Geiger classification, the whole territory is in the climatic zone C (temperate) (Fig. 3). Contextualizing this classification in the territory, according to Euskalmet (2023), three climatic zones can be distinguished: the temperate oceanic climate of the Atlantic slope to the north, very rainy and moderate in temperatures; then, a transition zone in Álava; and, finally, in the extreme south, a Mediterranean climate characterized by a clearly dry and hot summer of continental type, with a winter of low rainfall.

Fig. 3
figure 3

Koppen climate classification. C: temperate climate; f: no dry season; s: dry summer; a: hot summer; b: warm summer. Own elaboration from (Instituto Geográfico Nacional, 2024b) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Climate and terrain conditions determine land use. Population centers are located in valley bottoms. The main corridor links Bilbao and San Sebastián, with lesser connectivity along the axes connecting these capitals to Vitoria-Gasteiz. Consequently, substantial differences in density exist across the territory (Fig. 4), with two major urban areas: the larger one around Bilbao’s metropolitan area and another around San Sebastián’s. Conversely, in the case of Álava, the capital itself has experienced growth, attracting the majority of the province’s population. In mountainous areas, there is a pattern of extensive and scattered agricultural settlement known as ”caserío” SPAFootnote 3, hamlet, with no clear population center. Nevertheless, in Álava, besides the capital, the population concentrates on small settlements.

Fig. 4
figure 4

Population density. Own elaboration from geoEuskadi (2023) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Silvicultural production covers many of the mountains in Gipuzkoa and Bizkaia and is characterized by fast-growing species such as pine. In Álava, agricultural plains and vineyards dominate in the Rioja Alavesa, interspersed by the Montes Alaveses, with extensive masses of Mediterranean oak and holm oak groves (Palacios-Agundez, 2011).

In this environment, the communication infrastructures in the Basque Country have developed historically encouraged and conditioned by its strategic location as a passage to the continent and by the need to transport goods between the main seaports and the central plateau (Meseta) in the context of the Iberian Peninsula (Castillo and Valdaliso, 2017). This location, together with the agricultural poverty of the maritime provinces and part of the interior, in addition to its status as a free-trade zone, and the availability of hydraulic, mining, and forestry resources, determined the specialization of the Basque economy in the activities of trade, transport, mining, and industry (Valdaliso Gago, 2013).

Linguistically, within the geographical scope of this study, two languages coexist with equal status: the Spanish language (Castilian) and the Basque language (Euskera). This matter directly affects toponymy both in its origins and its development. These languages belong to different lineages, creating a considerable distance between them (Sanchez, 1974). Additionally, their situation is asymmetrical insofar as, according to the UNESCO Atlas of the World’s Languages in Danger and the European Charter for Regional or Minority Languages, the Basque language is a marginalized and vulnerable language within its territory (Council of Europe, 1992).

Fig. 5
figure 5

Flowchart. Own elaboration

Throughout history, the peoples inhabiting the area have named and transformed the territory, adapting names previously in existence to the usage of their language at each moment, thus bearing witness to past linguistic realities. García Sánchez (2023) discusses the etymon’s original language and the toponym’s receiving language, ”corresponding to the most current linguistic stratum of the territory where they are found”. A significant portion of the toponymic corpus in the study area, especially in minor toponymy, is the Basque language (i.e., from the Euskera language). However, language policies developed in more recent history have influenced the names on official cartography, resulting in toponyms in both languages, and the distribution of languages is not geographically homogeneous. The competencies to provide official status to geographic names are complexly distributed, whereas in terms of standardization, the Basque Language Academy (Euskaltzaindia) (Euskaltzaindia, 2023) is the sole advisory board (Article 10 of Law 10/1982, of November 24, the fundamental law for the standardization of the use of Basque language).

Materials and methods

Following the principles of toponymic study methodologies with a geographical focus (Tort-Donada, 2022), attention will be given to the semantics, distribution, and interrelationships of place names. Typically, the structure of a place name is formed by a generic part that identifies the type of named geographical element and another specific part for each element, ”Calleja Albar” SPA (”White Pathway”). If the specific part alludes to another existing geographical entity, it is known as a referential toponym, for example, ”Farolako bidea” EUSFootnote 4 (”Lighthouse Pathway”). Our work addresses two types of classification: one based on this referential element, if it exists, and another based on the semantic content of the specific name. Finally, the relationship between the distribution of these names and different landscape characterization variables will be analyzed using GIS tools (Fig. 5). The two prepared classifications will offer different avenues for interpreting the relationship between toponymy and landscape.

The tools used were Extract, Transform, Load (ETL) and GIS tools; QGIS (www.qgis.org) in its version 3.30.0-s-Hertogenbosch was used as GIS tool, while FME software was used as ETL tool, specifically FME© Desktop 2021.2 (https://www.safe.com/fme/fme-desktop/) and, alternatively, the graphical process modeler of QGIS 3.30.0-s-Hertogenbosch itself.

Data sources on landscape

Spain ratified the ELC on November 26, 2007 (BOE of February 5, 2008), which entered into force on March 1, 2008. Similarly, the regional Basque Government agreed to adhere to this convention in 2009. To promote its development, five years later, it approved Decree 90/2014 on the protection, management, and planning of landscapes in the territorial planning of the Autonomous Community of the Basque Country as a tool to fulfill the objective of integrating landscapes into territorial planning. Therefore, with a well-established legal and administrative framework, both the definition of landscape and the instruments for its management are established, such as the Atlas de los Paisajes de España (Atlas of Spanish Landscapes) (Sanz Herráiz and Mata Olmo, 2003) at the national level and the Landscape Catalogs, Landscape Determinations, Landscape Action Plans, and Landscape Integration Studies for the autonomous community. The Atlas de los Paisajes de España is a study aimed at characterizing and identifying landscapes, providing an exhaustive classification of the territory by establishing a scalar taxonomy of landscape types. The project began in 1998 and was published in 2003. The last update of cartographic information dates back to August 2010. The regional Landscape Catalog of Euskadi also offers a homogeneous cartographic base in the study area for analysis. This tool includes evaluating and classifying landscapes in each functional area of the territory. Additionally, it integrates a set of descriptive chapters to identify landscape units, analyze their characteristics and values, as well as propose objectives for landscape quality for each unit. Furthermore, the geographical information materializes in a data set called ”Cartography of the draft of the Catalog of Unique and Outstanding Landscapes of the Autonomous Community of the Basque Country at a scale of 1:25,000 (year 2005)” (IKT and PAISAIA, 2023).

This cartographic work results in a mosaic of 5604 landscape units, comprising 64 landscape units within the Basque Country as a whole. For the present regional scope study, the taxonomic scale provided by the aforementioned Atlas de los Paisajes de España proves suitable. Within this study’s scope, this cartographic work presents a mosaic of 72 landscape units, grouped into 47 classes of landscape units, as depicted in the following Table 1 and illustrated in Fig. 6. In order to use a more compact list that facilitates an integrated interpretation of the data, a synthetic classification of 12 types of landscapes is alternatively chosen, as depicted in Table 2.

Table 1 Classification of landscape units
Fig. 6
figure 6

Distribution of landscape units, including their types. Own elaboration based on geographic information from the Atlas de los Paisajes de España (Sanz Herráiz and Mata Olmo, 2003) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Table 2 Classification of landscape unit types

Data sources on toponymy

Concerning official cartography and geographic information in the study area, Spanish legislation outlines a scheme of competencies distributed among the different levels of administration. These competencies are distributed according to scale: from smaller scales for national coverage projects to the highest degree of detail for local projects. The geographic information production activity is coordinated among the administrations through an action model called the Sistema Cartográfico Nacional (Spanish National Cartographic System) (González Matesanz et al., 2015). One of its tasks is the coordination between administrations in the treatment of official and standardized toponymy. Consequently, for the present study, the toponymic sources of the administrations that provide services in the territory are used: national, autonomic (regional), and provincial (Batlle, 2023). The coordination between local and regional entities also allows the results of local toponymic works to be reflected in the regional database, which facilitates the integration of local knowledge.

Therefore, the combination of different sources requires a process of data harmonization. This harmonization of study data serves a dual purpose: firstly, integrating the toponymy from various sources, and secondly, establishing a format that enables the maintenance of the toponymic corpus while reflecting the traceability of each element. Thus, a dataset containing essential information will be formed, preserving the link to the original information to retrieve additional data if necessary. For this purpose, the source will be encoded, field names will be standardized, and the date of data acquisition will be recorded. Geometrically, a point-based database will be chosen, comprising original geometry for point elements and extraction of centroids for linear or surface features.

National level: NGBE and First edition of the National Topographic Map (MTN) at a scale of 1:50,000

In the geographical area chosen for the study, the Autonomous Community of the Basque Country or Euskadi, the first MTN coverage, in this case at a 1:50,000 scale (MTN50), was published from 1929 to 1944 (Urteaga González, 2001). The density of road labeling seems to be conditioned by cartographic criteria and, in areas where human activity offers a greater number of representation elements (Rosselló i Verger, 2004), the graphic possibilities of labeling are fewer. In addition, the methodology used for the topographic survey also conditions the quality of the toponymic survey, as will be explained later. Finally, the roads are structuring elements of the planimetric survey itself.

The digitization process of analogic cartography undergone in order to obtain the National Topographic Base, Base Topográfica Nacional (BTN), specifically of the MTN in this case, cannot be ignored. Due to technical issues and difficulties of recognition, some of the names of linear elements were omitted in this process. To all intents and purposes, this means the loss of the toponym that has not been automatically recognized, which can only be corrected by manual digitization. Linear toponyms are especially sensitive in the process of digitizing and automatic recognition for the following reasons: the non-concatenation of characters of labels that develop along the geographic element; the orientation is not horizontal but is also arranged according to the geometry of the element; the filling with backgrounds and surface patterns of the map; and the impossibility of differentiating names of pathways with respect to water courses and other linear labels. After observing the failure in the automatic emptying of labels of linear elements, it has been decided to digitize the pathway labels manually. For all these reasons, both the odonymy of the BTN and the Nomenclátor Geográfico Básico de España (NGBE), Basic Geographical Gazetteer of Spain, and the set of digitized pathway names from the first edition of the map are handled in a combined way.

Regional level: Official Geographic Gazetteer of the Basque Country (NGO)

The toponymic corpus of the Basque Government acquired its current status through Decree 179/2019, dated November 19, regarding the standardization of institutional and administrative use of official languages in local institutions of the Basque Country under the European Charter for Regional or Minority Languages, ratified by Spain in 2001 (Boletín Oficial del País Vasco, Official Gazette of September 15, 2001). This regulation establishes the Official Geographical Gazetteer of the Autonomous Community of the Basque Country, Nomenclátor Geográfico Oficial de la Comunidad Autónoma del País Vasco (NGO), respecting competencies in linguistic normalization and the formalization of place names (Ugarte Garrido, 2021). This database originates from previous work collecting geographic names with a solid linguistic focus (Agirregoikoa, 2011). The gazetteer is linked to the harmonized topographic base at a 1:5000 scale, leveraging the geometric quality of this product for the study.

Fig. 7
figure 7

Representation of pathways which have associated odonyms in the toponymic corpus. On occasion, a road with two separate lanes may appear to be a single road with double the width. Own elaboration from Instituto Geográfico Nacional (2024a), geoEuskadi (2023), and provincial governments. Frame coordinates in km

Provincial Level. Historical Territories: Geographical information 1:5000 and foral cadastres

Due to the historical particularity in the distribution of competencies, the three historical territories (Álava, Bizkaia, and Gipuzkoa) have developed cartographic tasks independently (Gómez Piñeiro and Sáez García, 1991), but with a common working scale of 1:5000. There is no vocational toponymic database, but toponymy is part of the information contained in such cartography. In addition, taking into account that the three provincial councils have competencies in cadastre, we want to take into account the place names contained in this dataset since they have had a differentiated treatment and, as far as we know, have not served as a source for the creation of official gazetteers of any level of administration. In spite of seeking homogeneity in data coverage, it should be noted that, in the case of Bizkaia, the cadastral database lacks toponymic information.

Local Works

As outlined in the description of the study area, this territory has two co-official languages. Consequently, toponymy inevitably garners relevant interest both in the public sphere and within the administration, as well as among researchers. Hence, numerous local toponymic studies have been conducted (Mujika Ulazia, 2010), including examples of municipal or regional studies (Líbano Zumalacárregui, 2003; Sainz Echeverría, 1996; Elosegui Irazusta, 1969; Arbizu Varona, 2016). Since 2011, the regional government has promoted the creation of local toponymic maps following standardized procedures and characteristics (geoEuskadi, 2014). The content of these projects serves as the primary source for the review, updating, and enhancement of regional and local databases.

Selection of the odonymy in the complete toponymic corpus

The object of this study is not general but focuses on the place names of the roads and pathways, so-called odonyms. This sectorial approach allows us to target the interpretation in a more specialized way. Therefore, it is necessary to extract the entities representing traditional communication routes from the complete set of names, or in other words, the names of the roads must be selected. Given the heterogeneous nature of the sources, the catalog of entities is different in each dataset, making the use of ETL tools essential.

The combination of toponymic sources within the same territory involves analyzing the redundancy between databases. In this case, an automatic cross-check would be performed to mitigate this effect (Mitxelena-Hoyos and Amaro-Mellado, 2023). The graphical result of this selection of linear elements is shown in Fig. 7.

On the other hand, when the geographic information has undergone a higher degree of generalization, as is the case of national cartographic series, it is common to use the paraje site type, defining an extension of territory with similar characteristics, as a resource for unknown classification labels. Therefore, the existence of places that explicitly contain the term ”camino” SPA and ”bide” EUS (pathway) as a generic part of the name will be verified. This fact implies differentiating these lexemes from the lexicalized etymons, which are already part of the specific part of the name. For this purpose, the position of the particles searched for in the string of characters that make up the complete toponym is analyzed. This result must be supervised.

On another note, given that linguistic standardization of Basque onomastics is relatively recent, current cartography includes non-standardized terms and even spellings that represent the oral variant of names. For instance, in Basque toponyms, it can be found ”kamino” as a non-standardized graphical adaptation, or the lexeme ”kamio,” which is valid for this odonym extraction, albeit slightly different in meaning (Gorrotxategi Nieto, 2022). The semantic difference between the two lexemes, ”bide” and ”kamio,” lies in the former having a more general use, while the latter refers explicitly to a main road.

Classification of odonyms

Frequently, classification methods are closely related to each author’s conceptual framework. For instance, Perono Cacciafoco and Cavallaro (2023) outlined different classification proposals historically used from the mid-20th century to the present, based on naming strategies within semantic blocks. Another perspective is offered by Santos Mansilla et al. (2023), who argue that the classification also relies on the conceptualization of the territory by its inhabitants. Considering the numerous possibilities offered by toponymic classification, a combination of two different perspectives has been chosen: one based on referential elements and another on semantic classification. These two classifications will complement each other in interpreting the results.

Grouping of odonyms based on the type of reference element

The etiology of the names of the roads and pathways can be very varied: among others, names that refer to their use, their morphological characteristics, and related landscape elements. On occasions, when the name of the road itself refers to the places it links, whether it is the origin, destination, or even an intermediate element ”Camino Vecinal del Alto de Gainza por Lejar a Utarte-Berri” SPA (”Neighborhood Pathway from Alto de Gainza through Lejar to Utarte-Berri”), we can speak of a referential component in the name. In these cases, the road’s name serves as an indication of direction so that, although its onomastic value is questionable, they collaborate in forming the mental representation of the spatial information of this territory (Reszegi, 2020). These reference landmarks can belong to both physical geography and human geography. In the case of physical geography, they can be, among others, oronyms like mountain peaks or mountain passes and hydronyms like streams or springs. Regarding human geography, the most common instances are urban centers, but functional destinations are also found-places where agricultural tasks or industrial activities take place, among others.

After reviewing the dataset, the classification is designed to make it exhaustive, logically ordered, and discriminating in as much detail as possible. Hierarchical classification is drawn at two levels of disaggregation in order to facilitate subsequent analysis. The classification concepts are included in the Table 3:

Table 3 Classification of place names based on their referential element
Table 4 Semantic classification of odonyms

Grouping of the odonyms with semantic criteria

Onomasiology deals with the analysis of the lexicon of a language from its meanings. The semantic component of the name is related to the characterization of the landscape. In order to take this information into account, the transparency of the name is essential. Therefore, we must be able to understand its meaning, or, in other words, we need the semantic relationship between a lexeme and a reality in the language in which they are expressed to have reached our days. In the toponymic corpus analyzed, transparency is maintained in a large proportion. For this reason, it is even possible to identify dynamics in the landscape due to the lack of concordance between the semantic content of some names and the nominated reality. On occasions, it has been necessary to review the bibliography to fully understand borrowings from the past, traditions, or professions that have been lost. Language dictionaries (Lakarra et al., 2023; Real Academia Española, 2023) and geographical dictionaries (Departamento de Cultura, 2001; Rebolé Del Castillo, 2003) have also been consulted.

Table 5 Results of the classification of place names based on their referential element

The initial semantic classification was completed by taking into account other works with the same subject matter: (Atik et al., 2022; Sousa and Garcia Murillo, 2001). Subsequently, a reading of the database was carried out in order to adapt the semantic groups to the existing casuistry in the study area. On certain occasions, geographic contextualization is essential to understand the meaning, so the review is undertaken using GIS tools. Each semantic group is shaded so as to reflect the completeness of the classification, as shown in Table 4.

Combining placenames with landscape classification

Once all the original data has been processed, the information layers are combined to geometrically compare the distribution of toponymic data concerning the superficial layer information from the landscape catalog. The method used involves attribute combination by overlaying the layer of point-based toponymic elements onto the superficial layer of landscape classification using the GIS. Through this combination, the frequency of each combination can be calculated, forming a matrix of frequency distribution. Given that the classifications proposed for both landscape classification and toponymic classifications -referential and semantic- have different levels of disaggregation, two types of outputs are proposed according to this level, allowing for different types of result analysis.

Results and analysis

This section first presents the direct results of classifying the odonyms according to the established criteria. Next, the results of the combination of the two entries, toponymy and landscape, are presented. Once the original information has been processed, an overall descriptive analysis is carried out. Still, it is necessary to go into detail in situations of special interest. This interpretation is illustrated through examples.

Furthermore, in order to optimize the descriptive analysis, a suitable statistical index is required. Therefore, in addition to the frequency, the density of toponyms in each type of landscape is used. In this analysis, values obtained are related to the characteristics of each data set. In this way, the relationship between the characteristics of each landscape unit and the place names classified in that environment becomes evident.

Results and analysis of the referential classification

The referential classification gives information on which geographic element was highlighted as a reference to identify the access or passage route. This classification’s results are shown in Table 5 and in Fig. 8. Unquestionably, population settlements are an important focus, in addition to economic activities, which are always closely related to primary activity.

Fig. 8
figure 8

Referential classification of odonyms. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

With respect to the settlement, it is clear that the longer roads connect to population centers. Contrariwise, the more local pathways provide access to smaller settlements, or hamlets (caseríos SPA or baserriak EUS), which is the traditional form of dispersed settlement on the Cantabrian side of the study area. There are also a notable number of references to buildings of industrial activities such as weaving mills, mills, and forges, whose layout depends on the water courses, making them a destination for transport in two directions: the arrival of raw materials and manufacturing output.

Table 6 Semantic classification of odonyms
Fig. 9
figure 9

Semantic classification. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

Table 7 Distribution of odonyms, classified according to the referential element and by landscape units
Fig. 10
figure 10

Distribution of odonyms, classified according to the referential element, by landscape units. Own elaboration based on geographic information from the Atlas de los Paisajes de España (Sanz Herráiz and Mata Olmo, 2003) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Results and analysis of the semantic classification

The main difference between referential and semantic classification lies in the names of population settlements, where the reference is clear. Nevertheless, the name can have diverse origins: for example, a historical essence, political connotations, or connections with local lineages and onomastics. Therefore, these settlements have been excluded from the semantic classification, except for those that contain some explicit allusion to population settlements, such as ”Camino de las Chozas” SPA (”Pathway to the huts”) or ”Camino a los caseríos” SPA (”Pathway to the hamlets”). As a result, 1411 names have been excluded from semantic interpretation. The results of this classification are presented in both Table 6 and Fig. 9.

Compared to the rest of the categories, numerous elements coincide equally in both modes of classification, such as ”Putzueta bidea” EUS (Path of the wells); the reference and their semantic value are related to hydronyms. However, there are other cases where the semantic and referential essence differ. For instance, ”Camino Viejo de San Martín” SPA (”Old San Martin Pathway”) holds a religious reference but also a descriptive semantic aspect. Interestingly, in the section related to semantic classification linked to geology, there are numerous oronyms featuring characteristics of crags or rocky formations, often associated with geomineral exploitation, like ”Artxipi bidea” EUS (”Path of small rock”) referencing a quarry or ”Buztinzuri bidea” EUS (”Path of white clay”).

Results and analysis of the landscape-odonymy crossover

Once the frequency data for each type of name in the different landscapes are obtained, they are displayed in Table 7 for the landscape type aggregation level arranged for the classification of the name references. Descending to the level of disaggregation of landscape units, the matrix of results acquires such a volume that the representation employing a map is more comprehensible (Fig. 10). In the same way, Table 8 and Fig. 11 are arranged for the semantic classification.

Table 8 Distribution of odonyms, classified according to the semantics, by landscape units
Fig. 11
figure 11

Distribution of odonyms, classified according to the semantic value, by landscape units. Own elaboration based on geographic information from the Atlas de los Paisajes de España (Sanz Herráiz and Mata Olmo, 2003) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

In order to obtain comparable results and find the maximum intensity of occurrence of a semantic phenomenon in a specific type of landscape, the density of occurrence is calculated from the frequencies of the semantic classification. The reading of this Table 9 and its disaggregated version will allow us to focus on the cases in which the density is anomalous, either due to abundance or scarcity. Although this indicator is not dimensionless, it has been decided to approach the analysis by keeping the information for each volume.

Table 9 Odonyms density by landscape unit
Fig. 12
figure 12

Settlement type. Own elaboration from geoEuskadi (2023) and Instituto Geográfico Nacional (2024a). Frame coordinates in km

Overall considerations

Firstly, the observation of the data makes it unavoidable to focus on one of the sources of information used, which is the digitalization of the first edition of the MTN50. Out of the 1806 toponyms digitized from this historical cartography, 1222 correspond to populated places from where the pathway starts or to where it goes, which represents 68.7%. In contrast, in the rest of the databases, the proportion remains at 33% (426 of 1277). These places or urban centers can have different characteristics: small population centers formed by a group of houses around a church or hermitage, which is a common form of settlement in the Llanada Alavesa, or Mediterranean slope; or hamlets or groups of hamlets in areas of Gipuzkoa and Bizkaia, or Cantabrian slope (Fig. 12), with greater dispersion of the population.

Continuing with the oldest dataset analyzed (sheets published between 1929 and 1944) (Urteaga González, 2001), the distribution of these labels is not homogeneous throughout the territory. While in Álava, there are abundant labels referring to population settlements, this type of labeling is scarcely found in Gipuzkoa. The explanation for this difference lies in the methodology used to prepare this historical cartography. Whereas in Álava and Bizkaia, it was completed by classical topography, Gipuzkoa was surveyed by photogrammetry. The latter reduces the volume of field work and, therefore, the possibility of collecting toponymic information by means of oral surveys. These differences are illustrated in Fig. 13.

Fig. 13
figure 13

Irregular distribution of pathway names in MTN50. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

The cases in which these labels, which indicate destinations of population settlements, can provide a specific value are those in which the road route goes beyond two neighboring towns, linking more distant destinations and usually with greater importance, such as the head of the region or the reference city of the environment. These roads that structure counties are usually called ”Camino real...” SPA (”Royal Pathway...”) (Ursúa Irigoyen, 1990). According to all the toponymic sources used in the study area, there are five cases in Spanish (three of them lexicalized ”Caminorreal”) and one in Basque ”Erregebide” EUS (”Royal Pathway”).

Fig. 14
figure 14

Itinerary determination according to the name of the pathway. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

Going deeper into this concept of inter-county road, whether or not the qualification of ”Real” (Royal) is written on the map, the fact of keeping the name of a town far from the surrounding area on the label, in itself, qualifies the road itself. This is because it gives it a regional entity as opposed to radial or local roads that link contiguous towns.

There is also the case in which the label indicates not only the origin and destination but also a place of passage. This circumstance allows inferring that there are roads with identical origins and destinations but different itineraries (Fig. 14). Finally, as a reference to traditional long-distance roads, there are two cases of ”Erremoesbide” EUS (”Pilgrim’s Way”) that correspond to the Camino de Santiago (Gorrotxategi Nieto, 2022).

References to elements of physical geography appear more frequently in complicated orographic areas, with low population density, such as mountain passes, and environments in which there is also grazing and seasonal transhumance typical of the area. This group includes both proper names, like ”Camino del Gorbea” SPA (”Pathway to Gorbea”) and generic allusions to the orography, ”Camino del Vallejo” SPA (”Small Valley Pathway”). In the case of hydrography, there are 30 explicit allusions to springs and streams, which are referred to by their specific names. There are also references to elements related to the hydrological infrastructure, such as the ”Camino de la Noria” SPA (”Waterwheel Pathway”) or ”Presalde bidea” EUS (”Dam Pathway”).

In the section on economic activities, the names related to professions, like ”Camino de carboneros” SPA (”Pathway of Charcoal Burners”), have been semantically classified. There are also activities related to mining or mineral exploitation, such as ”Camino de la Calera”SPA (”Lime Kiln Pathway”), ”Arrobitxulo bidea” EUS (”Quarry Hole Path”) or ”Camino del Marmol”SPA (”Marble Pathway”). On the other hand, we have included the phytonyms that have traditional economic exploitation, such as ”Camino del Castañal” SPA (”Chestnut forest Pathway”), ”Camino del Prado” SPA (”Meadow Pathway”), or ”Pinu bidea” EUS (”Pine Pathway”). As for the agricultural activity, the toponym ”Usaegieta bidea” EUS (”Dovecote Pathway”), ”Camino de los corrales” SPA (”Pathway to the Corrals”) has been cataloged within this group. Functional destinations include indications to strategic destinations such as ”Camino al Sanatorio de Briñas” SPA (”Pathway to the Briñas Sanatorium”), or ”Plazaolako Trentxikiaren Bidea” EUS (”Pazaola Little Train’s Pathway”).

Considering the table of densities, we can observe that the landscape with the highest total number of labeled pathways is TP52 Ebro depression countryside (0.859 names per km2). Nevertheless, if the population names are omitted, the maximum density is gathered in the Cantabrian-Atlantic estuaries or bays, TP89 Cantabrian-Atlantic estuaries and bays (1.263 names per km2). This landscape has a large population density, both at present and in the past. However, there are comparatively few references to population settlements (20.75%), while there are many names related to economic issues (30.19%), among which those related to agroforestry activity (phytonymy 10.9%).

The lowest density of names of any semantic classification is reported in the type of landscape TP83 Upper Ebro canyons and gorges (0.175 names per km2). The explanation for the latter may lie in the low population, inaccessibility and remoteness of large populations, and the small size of the area, so it may be necessary to resort to local toponymic studies.

Furthermore, the next lowest density corresponds to TP90, of marinas and valleys of the Cantabrian coast (0.244 names per km2). Most of the names in this area refer to population settlements (148), with no semantic interpretation. In this territory, the most numerous semantic classification (37 names, 13.41% of the TP90) is that of pathway descriptors such as ”Bidezarretako bidea” EUS (”Old Pathway”), ”Erdiko bidea” EUS (”Middle Pathway”) or ”Camino Particular”SPA (”Private Pathway”).

Overall, as shown in Table 9, we have worked with a very high number of references with content related to population, 1643 out of 3072. The next most numerous class is that of economic activities. Disaggregating this class into the different economic activities, we find that agricultural and livestock activities are very numerous (538 features). Given the classified elements, it has been considered that in the phytonymy analyzed and in the names related to geology, the functionality and the economic activity component prevail over the descriptive one, and that is why the semantic classes CS0320- Geology (mineral) and CS0310- Phytonymy (agricultural, forestry, land use), have been included in the class of economic activities. To illustrate this decision, we can find examples such as ”Meatzetarako bidea”EUS (”Pathway to the Mines”), ”Camino de Cortabaso” MULFootnote 5, (”Pathway to the Forest of the Sheepfold”), ”Larrabide bidea” EUS (”Pasture Pathway”).

Fig. 15
figure 15

Distribution of communal woodlands, coinciding with the type of landscape TP11. CS0208, CS0310 y CS0311. Own elaboration from Instituto Geográfico Nacional (2024a). Frame coordinates in km

Comparatively, economic activities account for a third of the volume of names related to population settlements (CS01+CS00), as do semantic references to the road itself and related elements and activities, such as sales, bridges, and descriptors. We find examples of the latter such as ”Bideberri” EUS (”New Pathway”), ”Bidezarra” EUS (”Old Pathway”), ”Caminorreal” SPA (”Royal Pathway”) or ”Camino de la Cuesta” SPA (”Slope Pathway”).

The names related to physical geography are more numerous in TP11, with 62 out of 175 names in this category, highlighting the use of names related to oronymy. Examples include ”Camino de Valleoscuro” SPA (”Pathway of Dark Valley”), ”Lezeagako bidea” EUS (”Pathway of the Caves”), or ”Arbarakoate bidea” EUS (”Pathway of the Arbara Pass”). This territory has a lower population density, abundant forested areas utilized for timber harvesting, and mountainous pasture lands, encompassing the most extensive areas of communal forests (Fig. 15). In the Basque Country, communal lands bear witness to a traditional socio-economic organization, where the productive functions of these environments are essential (Bellosillo, 1988), especially traditional forestry and livestock uses. These lands have recently attracted interest for recreational, ecological, or cultural purposes. Consequently, there is a growing appeal and recognition of these areas (Alcázar González and Azcárate Luxán, 2000). In this environment, it becomes particularly relevant to pay attention to zoonymy (51, 9.2%) such as ”Otsaurte Bide Zidorra” EUS (related to wolf), ”Mandobidea” EUS (”Pathway of Mules”), ”Camino de las yeguas” SPA (”Pathway of Mares”) and ”Camino del guano” SPA (”Pathway of the Manure”, used as fertilizer). Additionally, phytonymy (24, 4.3%) plays a role with names like ”Camino Madero” SPA (”Timber Pathway”), ”La Dehesa bidea” MUL, (”Pasture Pathway”), and ”Senda la Ayera” SPA (”The Beech Forest Footpath”).

Discussion

This research work is based on the intimate relationship between toponymy and landscape, as well as on the link between these two disciplines and cartography, being the territory the pillar on which they are sustained. Specifically, we start from the premise that both toponymy and landscape are based on the perception of the inhabitants of a territory (Reszegi, 2020) in order to look for a correspondence between their geographical distribution. In this research, the usefulness of toponymy in describing and studying the territory is evident since it synthesizes the description of the places themselves in place names. In this context, geographic data processing using GIS tools is fundamental.

Cartography is the natural medium for the expression and representation of toponymy, but it imposes its own limitations, as evidenced in this research. Each administration addresses the approach to the territory from a different scale and degree of detail, which characterizes the geographic information they generate. One of the most visible examples of the conditioning factors of the cartographic support can be seen in the series of the first edition of the MTN50 of the IGN; the sheets produced by classic topography differ from those produced by photogrammetry in their toponymic content, both in quantity and quality. Therefore, it is proven that the cartographic support conditions the toponymic content offered by scale, graphic issues, or the different data acquisition methodologies used. The integration of toponymic sources from different administrations has mitigated the bias that the specific features of each dataset may contribute to the toponymic corpus formed for this work.

The geographical entities studied in this research, which compose the traditional pathway network, are structuring elements of the territory, which are highly relevant for its understanding. There are works on pathway toponymy (Alcázar González and Azcárate Luxán, 2005; Bellosillo, 1988) and on the importance of pathways in the territory and landscape (Somoza Medina et al., 2022; Porcal Gonzalo, 2011; Orellana Moraleda, 2017; Rosete Vergés, 2020), also on landscape interpretation with the support of toponymy (Membrado Tena and Fansa, 2023; Nowosad and Stepinski, 2021; Sousa and Garcia Murillo, 2001; Spampinato et al., 2022). This work fits precisely at the intersection of these perspectives or disciplines and provides a systematic methodology for interpreting the landscape in a regional context. The success of this study lies in having identified common features in landscapes and place names based on a systematic classification of the latter, focusing on a specific geographical element, representing traditional communication routes. This is possible thanks to the functions of toponymy (Tort-Donada and López-Leiva, 2023): transparency, exceptionality, and territorial significance.

The work takes place in a bilingual environment that conditions the approach to dealing with toponymy, making proficiency in both official languages essential. Simultaneously, the complexity of multilingualism can serve as a stimulus for research in this field. In any case, it highlights the multicultural nature of the environment (Nyström, 2021).

Based on the landscape definition contained in the ELC, the intimate relationship between landscape and territory is understood; therefore, a landscape classification is a way of territorial classification based on landscape criteria. The classification of landscape units is founded on orographic concepts and land cover, where population and vegetation play a role, maintaining a visual and perceptual component as well (Marine, 2022).

Regarding the designed toponymic classification systems, while the referential classification focuses on key geographical elements for orienting within the network of roads, the semantic classification allows for a closer interpretation of the territory and the etymology of odonyms. As a trade-off, considering the complexity and diversity in the process of forming population names (Fort Cañellas, 1984), not always tied to the territory, it is preferable not to include this group of toponyms in the analysis systematically. The referential classification of place names allows for identifying traditional geographic focal points; it takes into account the toponyms that refer to or evoke geographical elements and classifies these references. The choice of these two classification methods has proven complementary in subsequent analyses.

As observed in the analysis of results, references to names linked to the population correspond to the distribution of settlement forms, scattered along the Cantabrian slope (except for more industrialized valley bottoms) and concentrated in small clusters forming a network along the Mediterranean slope. Agroforestry activities are present throughout the territory but are denser in mountainous areas, where traditional activities like forestry exploitation and high-altitude grazing are carried out, closely associated with communal land ownership, once again coinciding with specific types of landscapes.

Conclusions

In this work, two disciplines related to the perception of the territory, toponymy and landscape, have been combined, and the result confirms the coherence between them. The chosen geographical element for this study is the road network represented in official cartography, as traditional pathways have served as structuring elements of the landscape. Geographical contextualization of the data is crucial in this work, which is why GIS tools have been utilized. The analysis of odonyms concerning each type of landscape has evidenced the correspondence between the features of each landscape unit and the odonyms it contains. The distribution and semantic value of the toponyms have been used, along with their geographical referencing, where available. Therefore, the major innovation of this research lies in systematically intertwining disciplines that work with the inhabitants’ perception of the territory by analyzing elements that shape or structure it, such as the roads throughout the Basque Country Autonomous Community.

The level of detail observed in cartographic sources that encompass existing landscape classifications allows for opening new lines of research by applying this analysis to local geographical areas and considering various geographic elements of the territory. Additionally, the scale at which each study is conducted corresponds to levels of closeness to the territory and the breakdown of the mosaic of landscapes.

Another line of investigation, analogous to other geographical studies (Altuzarra et al., 2018) and landscape studies (Erikstad et al., 2022; Alcántara-Manzanares and Muñoz, 2017), is the study of the relationship between toponymy and landscape using tools like tools multivariate analysis (MVA) and principal component analysis (PCA). Regardless, our research validates the toponymy of a place as a tool for analyzing different aspects of its landscape.