Digging the Aboveground: Visual Archeology of an Asylum

Psychiatric practice during the nineteenth century was closely engrained in the spaces of the asylum. Envisioned as “therapeutic instruments” in and of themselves, asylums boomed all around the world: some were purpose-built, some were existing buildings repurposed as asylums. But how did these institutions really function? How was psychiatry practiced here?
Our proposal reports a work in progress that visualizes the evolution and workings of a nineteenth century Ottoman asylum. Located in Istanbul, the capital of the Ottoman Empire, it was used as the state mental hospital between 1873 and 1922. The building is originally a sixteenth century imperial complex by the famous Ottoman court architect Mimar Sinan (1489-1588) and is still standing albeit currently under heavy and damaging restoration. Research shows that during the time of its use as an asylum, the building changed significantly to enable its medical function. Indeed, it is precisely through the changing of the building one can demonstrate the medicalization and modernization of Ottoman psychiatry during this period. In this sense, our case study does not visualize an "ideal" asylum nor is it a search for the "original" complex; on the contrary, it analyzes a neglected and messy phase of an existing building's life. Traditionally perceived as a “corruption” in the original structure, we take the period it was used to treat the insane as a story worth being told.
The information is gathered from a detailed research of primary archival and printed sources involving multiple disciplines: medicine, psychiatry and architecture. These sources include: (1) Ottoman and Turkish official documents kept in the Prime Minister's Archives: Correspondences between state departments giving details about the running of mental institutions, spatial interventions, and concerns over public health; (2) Medical publications of the period: Visitor accounts by physicians, medical writings discussing approaches in psychiatry, and comparative reports on the current conditions of asylums around the world; (3) Popular media: Newspaper articles and travel accounts.
The project is located at the intersection of architectural and medical history; however, it is primarily using the toolset of architectural history. Partly stemming from the demands of the field and partly due to the nature of the project itself, one of the major concerns is the visualization of findings. In line with architects' particular understanding of visualization, the first resort was experimenting with various 2D and perspectival illustration techniques. 2D depictions (i.e. plan, façade, elevation) have been the mode of representation for centuries in the field of architecture. Perspective was a welcome addition with the development of linear, or mathematical, perspective in the fifteenth century. Computational 3D modeling and visual simulation (rendering) has become the norm starting from the 1990s. Despite its popularity in architectural practice and education, the use of 3D computation has, until recently, been relatively underutilized in architectural history.
Scholars in architectural history and in the neighbouring fields of art history and archeology have recently been using methodologies and approaches to integrate innovative technologies in historical research. From virtual reconstruction to web-based panoramas, geospatial modeling to photogrammetry, the virtual creation of static environments to interactive ones, these studies appear in multiple platforms and venues related to digital humanities, digital art history, architectural computation etc. In the mainstream publications of architectural history, an important milestone was the publication of an article that featured real-time interactive simulations of the Roman Forum presenting a hypothetical reconstruction of a funeral ceremony in the Journal of the Society of Architectural Historians (JSAH) (Favro and Johanson, 2010). Another outstanding project that addressed issues of time and movement was "The Virtual Monastery" that focused on a single structure to analyse the workings of a building type through ages (Bonde et al., 2009).

Visualizing the findings
In our project, we faced several challenges to demonstrate the findings with methods traditionally used in architectural history. Ottoman mental facilities in general, and our case study in particular (especially the period that we are focusing on), is largely understudied compared to the aforementioned examples of the Roman Forum or monasteries. All the findings come from primary sources that are incomplete in nature. Moreover, the sources are predominantly textual with limited visual resources describing various states of the building. The historical layers added to the building render conventional visualization methods ineffectual. As a response to these challenges and the resulting complicated presentation of findings, the project employs an interactive 3D simulation toolset built within Unity 3D that allows visualizing hypothetical spatial reconstructions and trajectory tracking all continuously referring to temporal data and incorporating primary documents. Our interactive simulation framework builds on these efforts creating the ability to navigate in four dimensions simultaneously and without restriction:

1 Temporal dimension: Instead of the traditional interest in finding out about the “original” building, this approach looks at the
life of a building and how it changed over time. The method we propose for Unity is to implement a time slider component into the viewing user interface (UI). This component controls the visibility of architectural changes at distinct time periods. In this project, we have identified four specific points in time wherein primary sources identify distinct architectural states. The timeline UI is enabled by including time data within the model data. In the 3D modeling or 3D data acquisition phase, model elements are tagged with custom properties coded to our tool. These custom properties represent abstract metadata defining the start and end date for the object(s) (for temporal properties), the orientation point for the object(s) (for geospatial locating), as well as any descriptive information all of which become native to the file.

2 Reconstructed trajectories: Instead of treating space as an isolated entity, this approach captures the
life in the building. By tracing the movement of the occupiers of the space, we know more about their lives. The everyday life in the asylum consisted of the acts and movements of its occupiers, in this case, patients, doctors and the staff. By tracing their movements and visualizing them, it becomes possible to have a better understanding of the daily routines (eating, cleaning, sleeping etc.) in addition to the medical treatments that took place through a certain reorganization of space and time. This setting is particularly illuminating as it was assumed at the time that the life and routines of patients in the asylum were crucial components of the healing process.The user interface (UI) enables diagrammatic circulation indicators for each of the four points in time to be visible at the viewer’s discretion.

3 Cross-referencing data: In addition to the ability to preview time and movement, the interactive functionality in Unity allows the viewer to select the information relevant to the simulation viewed at any time. The ability to combine text as hyperlinked visual overlays and graphic information enables important reference texts and images to be associated with spatial and temporal elements and viewed when desired. One advantage of this presentation method is the ability to break free from the linear structure of a textual manuscript. One can interact with the information presented in a nonlinear way (albeit guided). By cross-referencing various types of information, we aim to create a system that binds or connects these documents and data sets in meaningful ways.

Figure 1: Simulation User Interface Mockup (B. Ozludil, U. Thompson, A. Wendell). Photograph source: Sihhat Almanaki, 1933.

We do not intend the spatiotemporal model to be the
primary outcome, but rather a platform inviting interpretation and scholarship through the combined textual and visual data. In line with the ontological approach adopted in digital humanities, the ever-evolving outcome is a “knowledge representation, a hypothesis” not a re-creation of reality (Bonde et al., 2009). The relationship between the representation and its referent also calls into question producing knowledge with partial data, which is not uncommon in (conventional or digital) historical scholarship. Both possibilities and risks exist in 3D visualization with partial data. 3D models give the opportunity to relate to the building without the trained eye to "read" 2D representations. Risk is related: 3D creates a sense of absolute masking the hypothetical elements within. Being aware of this danger, the project aims to visually clarify the “known” versus “unknown” using graphic coloration indicating the extent of “known” detail.

The tool we propose visualizes temporally-located spatial data that allows multiple readings and interpretations, and that is open to manipulation, rather than a mere representation of a space. Doing so will open the approach to more disciplines and interested parties, and while still technical in nature, will provide a platform for historical spatial research. This work can be shared with other digital humanities scholars employing Unity or similar tools. Integrating innovative technologies in historical research has the potential to change the ways in which we conceptualize and tackle the problem at hand. In other words, rather than being "tools" to accomplish what is predetermined, they open up new ways to think and to produce knowledge.

Bibliography

Alkhoven, P. (1991). The Reconstruction of the Past: The Application of New Techniques for Visualization and Research in Architectural History. Computer Aided Architectural Design Futures: Education, Research, Applications. Zürich (Switzerland), pp. 549-566.

Boeykens, S. and Neuckermans, H. (2009). Architectural design analysis, historical reconstruction and structured archival using 3D models: Techniques, methodology and long term preservation of digital models. In Tidafi, T. and Dorta, T. (Eds.) Joining Languages, Cultures and Visions: CAADFutures 2009, PUM, pp. 119-32.

Bonde, S., Clark M., Elli M. and Julia F. (2009). The Virtual Monastery: Re-Presenting Time, Human Movement, and Uncertainty at Saint-Jean-des-Vignes, Soissons. Visual Resources: An International Journal of Documentation, 25(4): 363–77.

Bourdakis, V. and Pentazou, I. (2012). Real City Museum/Virtual City Model: Real Datasets/Virtual Interactions. In Achten, H., Pavlicek, J., Hulin, J. and Matejovska, D. (Eds.), Digital Physicality - Proceedings of the 30th eCAADe Conference, 2, pp. 337-41.

Drucker, J. (2013). Is There a "Digital" Art History?. Visual Resources: An International Journal of Documentation, 29 (1-2): 5-13.

Favro, D. and Johanson, C. (2010). Death in Motion: Funeral Processions in the Roman Forum. Journal of the Society of Architectural Historians, 69(1): 12-37.

Gill, A. A. (2009). Digitizing the Past: Charting New Courses in the Modeling of Virtual Landscapes. Visual Resources: An International Journal of Documentation, 25(4): 313-32.

Kalay, Y., Kinayoglu, G. and Kim, S. W. (2005). Spatio-Temporally Navigable Representation and Communication of Urban Cultural Heritage. Proceedings: VSMM 2005 International Conference on Virtual Systems and Multimedia. Ghent, Belgium, pp. 145-52.

Osman, M. (1933). Sihhat Almanaki. Kader Matbaasi, Istanbul.