Incomplete map of routes covered by research team members; total about 20,000 km, October 2018 – October 2024.  Map by Dr. Ivan Marić

We designed the research to overcome some limitations in the existing literatures on B&H that were written during or after the 1990s war.  First, we did research throughout all of Bosnia and Herzegovina (see map).  This meant that we investigated the changing religoscapes of all communities, not only of Bosniaks, or of Serbs, or of Croats, but also of the few built by other religious communities.  We also paid attention to the configurations of settlements and the ways that the boundaries of ethno-territories were marked.

Further, most studies on the damage or destruction of “cultural heritage” take an ahistorical view of the religioscapes existing at the start of 1992, as structures on an otherwise neutral landscape in which specific actors engaged in the criminal destruction of Bosnian national heritage.  Yet the buildings had not been randomly scattered, and neither were their features necessarily only the result of parochial tastes when they were built.  The theoretical model that we employed, derived from the Antagonistic Tolerance project, analyses specific buildings, sites and configurations of religioscapes at any point in time as reflecting the trajectories of social relations between the communities building them that had existed until then, and when possible we also analyze later trajectories of events.  

Accordingly, the research team engaged in fieldwork, on three scales:

Survey throughout B&H: We documented locations of buildings, cemeteries and other relevant features and photographed them with geotags so that they could be entered onto maps.  The Antagonistic Tolerance indicators of dominance are centrality in a settlement (at varying scales) and perceptibility: features that heighten the visibility of a site (e.g. height, color, sheer massive size) or audibility (bells, call to prayer); some features may serve both purposes.  Religoscapes and the physical attributes of the structures comprising them were evaluated by site visits analyzing the features of buildings themselves, and in their relationship to each other.

From the earliest field trips it became clear that the locations of memorials to people killed in the 1992-95 war were the most critical indicators of which community was dominant in any location.  However, there were also such monuments in settlements that had been emptied or nearly so, in which case they represented the presence of the absent.  In both kinds of case, with almost no exceptions, such monuments were linked to only one of the ethno-religious communities of BiH, with corresponding wording and imagery.   That is, monuments dedicated to killed soldiers of the Army of Bosnia & Herzegovina (ARBH) use Koranic quotations, while those to the dead of the Croatian Defence Council employed Roman Catholic visual imagery utilized.  Monuments to the dead of the  Bosnian Serb Army (VRS, Vojska Republike Srpske) are written in the Cyrillic script and usual visual imagery of Orthodox Christianity. 

Case Studies of Changing Religioscapes in Five Settlements, 1862-2022.

We tested the model developed in our survey research in case studies in which we could read the changing religoscapes from 1862, when the Ottomans had built five new settlements on the Una and Sava rivers to house Muslims expelled from Serbia.  Thus we have a starting point of the structures of a purely Muslim settlement.   As political dominance shifted, changes in the physical layouts of the settlements changed, and churches were added and substracted to the religioscapes from the 1870s until after the war ended in 1995.  In fact, the most sustained period of religoscape construction after 1862-63 was after 1995, though differently in the five towns.  This work was based on historical and demographic sources as well as site visits.

Case Studies of the Religioscapes of Cites:  Banja Luka, Mostar and Zenica, 15^th century to 2022

We next tested the utility of our model on three cities that also began with small Ottoman base layers but developed over the centuries into large, socialist industrial cities, which thus added socialist secularscapes to our framework.  Zenica, is overwhelmingly Bosniak after the 1990s war, while Banja Luka is overwhelmingly Serb, and Mostar is de facto divided between Bosniaks and Croats.   We brought geographers and historians into the fieldwork.  These studies used innovative GIS mapping techniques to demonstrate the sedimentations and erosions through time of religious structures and monuments in the three cities.   Detailed studies of each city were published in 2024 in a thematic issue of the journal Geoadria: https://hrcak.srce.hr/broj/25324.  The Introduction to the thematic issue has much more detail about the methodologies and theories utilized: https://hrcak.srce.hr/330381 .  

In the special issue of Geoadria (Vol. 29 No. 2 [2024]) on “Spatializing Sedimentations and Erosions of Time in Urban Landscapes” focused on the development of Banja Luka, Mostar and Zenica, (https://morepress.unizd.hr/journals/index.php/geoadria/issue/view/403)

we applied the Antagonistic Tolerance model to cultural geographic data to analyze manifestations of changing local dominance through the sedimentation and erosion of sacral sites.  The methodology workflow was divided into five main steps: (1) the acquisition of historical Austro-Hungarian (AH) maps of selected cities; (2) the harmonization of AH historical maps; (3) the derivation of a GIS database (4) the acquisition of attributes on "erosion" and "sedimentation" of religious buildings; and (5) visualization, i.e. maps derivation.

1.1. Acquisition of historical Austro-Hungarian (AH) maps

Historical AH maps of Mostar, Zenica, and Banja Luka served as a template for mapping the initial or reference state of religious buildings, from which their "erosion" and “sedimentation" would be monitored over time. Historical maps of BiH were made during the Third Military Survey (1869-1887), with a scale of 1:25000. They were selected as maps of the "reference state," i.e. the starting point of the analysis. The maps of the mentioned cities were downloaded from the Arcanum Maps (formerly Mapire) geoportal using the printscreen option.

Arcanum Maps is an online publisher that produces structured databases of digitized cultural content (maps, newspapers, etc.). The printscreen download was made to cover the entire area of ​​today's settlement boundaries of the three selected cities. While print screening the images, front and side photo overlap of approximately 80% was maintained to solve the problem of relative orientation and, ultimately, their georeferencing (absolute orientation) in the defined global coordinate system WGS 84/UTM zone 33N—EPSG:32633 of the default GIS database.

1.2. Harmonization of historical AH maps

The problem of the relative orientation of the collected historical maps was solved using the Auto registration option within the Georeferencing tool (ArcMap 10.8.1). Auto registration automatically allows repositioning a specific raster to a referenced raster dataset. The automated links (Figure 1) are created based on spectral signatures (pixel colour). Auto registration of acquired screenshots was done image by image. For example, around 30 screenshots were collected for the city of Mostar. Therefore, the problem of relative orientation using the auto registration option was solved for screenshot 1 and screenshot 2. This created a new raster dataset, Image12. Then, relative orientation was solved for the newly created raster Image12 and screenshot3. Following this procedure, the problem of screenshot merging was gradually solved until a fully merged historical image of Mostar was obtained, which includes all 30 acquired screenshots. This procedure was performed for all three analyzed cities.

Figure 1. Example of solving the problem of relative orientation

After relative orientation, merged historical images of the analyzed cities were geo-referenced; they were harmonized or aligned with today's satellite images using marked fixed control points. Georeferencing refers to the methodological procedure by which historical maps are harmonized with the modern coordinate system to enable their GIS application, i.e. the possibility of performing different spatiotemporal analyses (Figure 2). The georeferencing process of merging historical images for each city was done using the Georeferencing tool's Add Control Point option. For control fixed points, fixed objects that did not change their location during the studied period (time of making the map - today) were selected. About twenty fixed points were used for georeferencing within each city.

Figure 2. Example of a georeferenced historical maps for the settlement of Zenica

1.3 Derivation of a GIS database

Georeferenced historical AH maps of selected cities were saved in a GIS database (format: file geodatabase). Additional raster and vector data needed for further analysis and visualization were stored in the same database: satellite images of selected cities, current traffic and residential infrastructure (Geofabrik, 2024), traffic and residential infrastructure in 1887 (manual vectorization), urban area of ​​settlements during AH (manual vectorization) and water flows (1887 and 2024). A topological check was performed on all downloaded and acquired vector data.

1.4 Acquistion of attributes on “erosion” and “sedimentation“

For the analysis of spatiotemporal changes in religious buildings, that is, "erosion" and "sedimentation," it was necessary to acquire the corresponding attributes of the buildings' state. Therefore, a table template was created and filled in with data from various sources (field research, literature analysis, interviews) (Table A).

The table contains the following data: ID (specific code of the religious object), year of construction, notes (specificity of the object) and object condition in a particular year. The condition of the religious object is described by the following legend:

(0) the object did not exist in that year;

(1) the object exists, has its function and is not damaged;

(2) the object exists and is damaged

(3) the object is significantly damaged or completely demolished

(4) the object exists but has completely lost its function

(5) renovation of the building has started or is under construction.

Table A. Template of table used for data acquisition

Considering the results of preliminary analyses and the fact that the period of spatiotemporal analysis is more than 130 years (from 1887 to 2024), it was necessary to define selected years for which accurate data on the condition of religious buildings would be collected. It was decided that the state of religious buildings would be mapped for four years: 1888, 1961, 1991-1995. and 2024. The first year (1888) refers to the time immediately after creating the map used to map the reference state. In other words, it maps the religious objects that the AU administration recorded immediately after the occupation of Bosnia and Herzegovina in 1878. The second year (1961) depicts the state of religious buildings in the Socialist Federal Republic of Yugoslavia (SFRJ). For the third year, the period (1991-1995) was used to depict the state of religious buildings during the war in BiH, fought between September 1991 and December1995. The final fourth year (2024) reflected the state of religious buildings three decades after the end of the war.

After deriving a table template and selecting reference years, data acquisition on "erosion" and "sedimentation" of religious buildings began. The data was collected through field research, interviews, literature analysis and various internet sources. Field research included visits to Mostar, Banja Luka and Zenica to record religious buildings directly. Namely, after all, religious objects were mapped in GIS on the georeferenced AH map template, a field survey was carried out, and all current religious objects in each city were added to the GIS database. Furthermore, each object was photographed, a specific ID was added, and a note was added if necessary. All mapped objects were then checked using the Google Earth program.

1.5 Data visualization

The last step of the methodology refers to the visualization of data, that is, the creation of thematic maps that will depict the intertemporal patterns of sacral architecture of selected cities. Given that the state of religious buildings was mapped for the selected four years, the map of each city consisted of four data frames. A georeferenced AH map was used as a basis for the oldest year, while a satellite image was used for more recent periods.