Abstract
Early modern Istanbul, the Ottoman Empire’s capital, was a major urban center well-known for its large religious complexes and bazaars, a masonry architecture of lead-covered domes still enduring. In contrast, the residential architecture was mostly timber-framed, and due to the material ephemerality in fires, little is known about and left of the city’s residential neighborhoods. This article considers the Phanar waterfront—the neighborhood of a group of Greek Orthodox Christians known as Phanariots—as a significant exception to this urban historiographical enigma. The article visualizes the data on Phanar semi-masonry houses in historical sources with digital three-dimensional (3D) tools as it documents, reassembles, and investigates the neighborhood’s main street. Taking the mid-nineteenth century as a hallmark of reliable representation and diachronic analysis, the article aims to demonstrate digital 3D reconstruction as a scholarly method for documenting and researching lost urban heritage.
Keywords
In his 1852 Istanbul travelog, Théophile Gautier, the celebrated French writer, narrated an excursion through Phanar, one of the city’s Greek Orthodox Christian quarters, located on the southern Golden Horn between Balat and the city’s main harbor. He pictured a striking contrast between “the wretched Balat” neighborhood, a major Jewish settlement, and that of Phanaris, of “a better class of Greeks”: “a sort of ‘West End’ in proximity to an ‘Alsatia’ or a ‘Cour des Miracles.’” 1 What made Phanar comparable to London’s high-class neighborhood while Balat matched Alsace’s towns with timber-framed buildings or Paris’s old slum district was rather formal and related to residential architecture. Phanar consisted “of houses of stone, of considerable architectural pretension.” Some were more ancient and recalled the French mansions of the Middle Ages: “half-fortresses and half-dwellings.” After a long poetic description of the house façades, he signified the one-mile-long high street between the fortifications and harbor as “the masonry neighborhood.” 2 Istanbul, which Gautier visited, was, as the Ottoman reformative government had imagined and tried to promote, at the beginning of a long process of urban transformation from a city of timber residential neighborhoods to one with masonry buildings. Declaration of building codes, the realization of post-fire partial plans, state offices with stone façades, and the competition among privileged and ruling classes to have masonry residences were steps toward this vision. 3 Within a slowly changing city, however, the Phanar masonry neighborhood was still significant because it was an ancient and early modern quarter, and its masonry houses were exceptional examples of a time, from the mid-seventeenth to early nineteenth century, when most of Istanbul’s residential buildings were timber-framed. 4
Phanaris, better known as the Phanariots, were Greek-speaking Orthodox Christian mercantile elites, some of whom would rise within the Ottoman state apparatus as imperial dragomans and as appointed rulers (hospodars) of vassal principalities of Moldavia and Wallachia (mostly in present-day Romania) forming a complex power network covering the whole Ottoman Empire and beyond. 5 After the Greek Orthodox Patriarchate’s move to the Petrion Castle by the Phanar Gate in the 1600s, these elites, including the Wallachian and Moldavian aristocrats, started investing in the extramural waterfront by building semi-masonry residences, replacing the old fisher houses there, and expanding their plots toward the sea. 6 Eventually, a masonry house in Phanar evolved as the Greek Orthodox elites’ prerequisite enterprise for aiming to have privileged positions within the Ottoman State and the Danubian Principalities. The Phanar waterfront was unique in Istanbul and emerged in one-and-a-half centuries as the collective creation of an ambitious and competitive community, who would peculiarly be named after their neighborhood. 7
Before the nineteenth century, the Phanariots’ neighborhood might have constituted the largest single area of masonry residential buildings outside Galata and Pera. Following the end of the Phanariot ascendency and statesmanship from 1821 to 1850s, these houses started to decay and be replaced by new structures slowly until the late twentieth century. At present, there are only a handful left in situ, the majority of which have been recently converted to exhibition venues by the Istanbul Greater Municipality. 8 Fortunately, when it was slowly transformed and demolished, the Phanar masonry neighborhood was documented by the city authorities, antiquarians, art historians, and travelers. Defining the mid-nineteenth century as a hallmark and making a diachronic analysis of successive visual sources, this paper aims to reassemble Phanariots’ masonry neighborhood.
Urban historical representations of early modern Istanbul residential neighborhoods, devasted regularly by fires being timber-framed constructions and continuously rebuilt until the mid-nineteenth century, are mostly textual except for some Galata neighborhoods and Bosphorus suburban villages. 9 In the special example of the Phanar masonry neighborhood, this paper presents a rare attempt to make visual research of a peculiar early modern Istanbul residential quarter in detail as it survived to the mid-nineteenth century. 10 Based on a three-dimensional (3D) visualization research project, 11 it aims to contribute to the architectural and urban history of Istanbul, the Ottoman Empire, and the early modern Mediterranean by documenting and representing the now widely lost built and urban heritage of the Phanar waterfront houses. By embracing a source-based approach, the historical neighborhood has been digitally reconstructed by installing and integrating textual and visual data from historical sources onto a 3D model. As the diversity in the content, scope, origin, format, and detailedness of historical sources challenge the applicability of data, the complexity in the spatialization of historical evidence has been addressed by three methodologically reciprocal modalities to digitally document the disappearance of and reassemble the historical urban fabric of Phanar: (1) diachronic analysis, (2) base map production, and (3) 3D modeling. Thus, an equivalent contribution of this paper is to demonstrate the potential and limits of 3D modeling in urban historical research as a scholarly method.
The paper starts by introducing research by visualization and source-based methodology adopted in digital reconstruction as positions for reassembling the Phanar masonry neighborhood. The following part documents the primary visual sources by narrating the history of Phanar houses’ disappearance and the present-day in situ data. Then, the paper unfolds the research process by illustrating different stages of analysis, data application, modeling, and representation and presents the two-dimensional (2D) and 3D outcomes of the visualization project. The concluding parts offer possible outcomes and further lines of research discovered in Phanar’s masonry neighborhood’s visualization and discuss the methodological challenges, critical tools, and implications of a lost historical urban fabric’s modeling and visualization.
Visualization by Digital Tools as Historical Research and Source-Based Methodology
Digital humanities in Ottoman historiography is emerging. 12 The extensive textual archives of the empire provide an opportunity to experiment in researching with digital tools. Most of these studies use artificial intelligence and informatics to handle big data. There are also studies focusing on information mapping by geographical information systems (GIS). Some other architectural and urban history research utilizes digital tools for visualization, not only for representational but also for historiographical purposes. Most of these combine conventional architectural graphic communication techniques with digital 3D modeling. While the historical information on the Phanariots, the Phanar network, and Phanar is very diverse and suitable for all these digital humanities approaches, the last approach, research by 3D visualization, is adopted in this article. The precedents of this approach from urban history studies will be briefly summarized to clarify the methods applied here.
The idea of representing change through time—the fourth dimension—by 3D modeling to visualize and thus analyze the transformations in the built environment is at the core of many research initiatives. The research projects conducted by Wired! Digital Art History and Visual Culture Research Lab at Duke University, particularly the collaborative project of Visualizing Venice, provides a significant precedent for mapping, 3D modeling, and representing change over time by visualizing the city’s historic fabric. 13 The Visualizing Venice project defined visualization as fundamental to constructing an “embodied engagement with spatiality.” It utilized digital modeling as a new level of concreteness in historical research and as a novel form of cognition and understanding of historic buildings or sites. 14 The Visualizing Venice project emphasizes “visualization as a research method” by conducting a discovery-oriented process using digital tools for analyzing and exploring historical documents in new ways. Among other inspirations and precedents for this paper is the four-dimensional (4D) Research Lab at the University of Amsterdam, where digital 3D reconstructions of historical, archeological, and architectural settings are used as research tools. 15 The Freedom of the Streets: Gender and Urban Space in Eurasia (1600-1850) 16 is a research program led by the University of Amsterdam that analyzes gendering urban space in the early modern city in four closely related research projects with diverse methods focusing on specific cities or themes. The project “Digital Urban History of the Early Modern World: Amsterdam and Edo (present-day Tokyo)” by Gamze Saygı, Bob Bierik, and Marie Yasunaga is significant as a reference to streetscape history since it utilized digital methods to explore the spatial history of these two cities through 3D reconstructions and story maps of their street lives. 17 HistStadt4D at TU Dresden is another research project focused on the representation of urban history, which establishes a 4D perception of Dresden’s transformation by correlating historical sources, by georeferencing and spatially orienting historic photographs particularly, with a virtual model. 18 Regarding the studies on the visualization of architectural and urban history in the context of Turkey, Ali and Müge Cengizkan’s exhibition and book on early Republican Ankara as the capital city in 2019 titled The Construction of a New City: Ankara 1923-1933, which includes a digital reconstruction of “Yenişehir” (New City), presents a valuable precedent. 19
These research projects demonstrate the diverse use of 3D modeling in digital humanities as the basis of analysis, interpretation, and reconstruction of transformed or disappeared historical buildings and sites. Contributing to these endeavors in establishing digital 3D reconstruction as a scholarly method in architectural and urban history research, 20 the research project that this article is based on has implemented a “source-based methodology” to investigate the disappeared urban fabric of Phanar’s extramural waterfront through digital 3D reconstruction. The 3D modeling of the historical urban fabric was predominantly based on historical sources, namely, historical maps, city views, panoramas, architectural documentations and drawings, and photographs, which are supported and corroborated by in situ surveys and up-to-date cadastral plans. As the research rigorously collects, installs, and correlates these various types of historical evidence to support the understanding and research of the disappeared urban fabric, it produced an extended inventory of the Phanar waterfront houses. For the specific case of Phanar, this research project is one of the first to apply such a method. 21
However, it should be acknowledged that the challenges of implementing source-based methodology in 3D historical reconstruction and visualization are manifold. In an example like the Phanar waterfront that has been drastically changed through the twentieth century, and even though there are singular houses still standing on the waterfront, the physical traces of the distinctive urban fabric in situ have vanished, it is not possible to conduct a reality-based 3D documentation of the case location, such as by laser scanning. While the persistent monumental structures, such as religious buildings and city fortifications, which earlier studies had studied, are evaluated and used as points of continuity between different periods, the inconsistencies in the availability of historical evidence on waterfront houses exacerbate the scarcity of extant physical evidence for urban-scale digital reconstruction. As the availability of historical evidence varies, the granularity of data (i.e., the level of detail and precision of data) collected for different historical periods, urban fragments, and houses fluctuates between non-scaled traces of building footprints to measured detailed drawings of a house’s interior.
All these challenges of producing a source-based digital reconstruction necessitate an interpretive process of balancing and correlating the data in different types and levels of detail to integrate historical evidence in architectural and urban scales. Modeling in historical visualization is fundamentally relational and evidence-based, while it designates a heuristic process, not a denotative act of representation. Digital modeling (re)constructs the world it represents as it alters how we think about it. Caroline Bruzelius identifies digital work as a “modus concipiendi,” which provides novel ways of thinking and questioning historic buildings or sites that are transformed or disappeared as it generates new knowledge by fusing evidence. 22 Likewise, this paper approaches digital reconstruction by 3D modeling as a research tool to implement a scholarly method for documenting and investigating the early modern houses on Phanar’s extramural waterfront. It embraces the digital 3D model as a means of developing a hypothesis through which new questions arise regarding the architectural and urban distinctiveness of the historical context, while various historical questions about the urban morphology of Phanar’s extramural waterfront, the patterns of ownership, the particularities of the Phanariot’s houses as an architectural typology and the lived space can be dissolved.
The source-based digital 3D reconstruction provides virtual and visual access to a disappeared urban fabric that might otherwise remain inaccessible. However, the representational primacy and the visual power of the 3D model should be critically addressed in historical research. When established as the ultimate result to be achieved, the digital 3D reconstruction replaces the lost past by making the absent present with a virtual replica of the historical context. The historical visualization here is neither an artistic representation nor an aesthetic model; it does not intend to provide a photo-realistic result and avoids a polished visual image. Peter Burke claims that the use of images as historical evidence and period displays based on historical reconstructions can give “a sense of direct contact with life in the past,” 23 which Roland Barthes identifies as the “reality effect” 24 evoked by documentative and archival sources such as historic photographs. Yet, Burke reminds us that such “immediacy is an illusion.” 25 Avoiding a harsh effect of reality or an illusion of actuality, the digital model does not aim to construct an urban digital twin or a mirror world of Phanar, 26 it is the tool and the medium for documenting, interpreting, and communicating research data to generate new knowledge.
Slow Disappearance of the Phanar Masonry Neighborhood
Around the 1810s, there were over 130 houses on the larger Phanar waterfront, as listed in Head Gardener registers (Bostancıbaşı Defterleri). 27 Some of these were confiscated after 1821 at the start of the Greek War of Independence, when the Phanariots were declared complicit, had fallen communally from grace, and were punished by the Ottoman State. Some other houses were deserted as their residents escaped from Istanbul. In a decade-long process, Phanar would be restored partially to its former owners and grandeur. This was the Phanar of the 1830s and 1840s, the time of neo-Phanariots such as Stephanos Vogorides (Prince of Samos) and British traveler Miss Pardoe, who visited the waterfront houses attending some festivities and banquets. 28 However, in the mid-nineteenth century, during Gautier’s visit, the community that made Phanar was once more on the move, this time voluntarily to the newly constructed masonry quarters of Istanbul, majorly Pera. Few would remain and preserve their heritage, while new residents would inhabit the remaining houses and alter them. This section constitutes a survey of the available visual sources on the Phanar waterfront from 1850 to the following 170 years until the present day by narrating the long disappearance history of the masonry neighborhood of the Phanariots: How the greater Phanar waterfront was transformed after the 1850s as part of Istanbul’s successive urban transformations; how was it rediscovered art and architectural history wise and documented in the point of its destruction?
Kargirs of Phanar: Traces of the Masonry Neighborhood from 1851
In 1851, one year before Gautier’s tour, Greek writer Skarlatos Byzantios had published the first of a three-volume book series on Istanbul’s historical topography, which also covered Phanar. 29 There, he mentioned the Christian “seaside mansions made of stone,” specifying them as kargir (γκεργκύρι). Kârgir means “a building built by stone or brick structure; masonry” in Turkish (from Persian) and had been adopted in Greek for a masonry seaside mansion. This naming from the Greek Istanbulite community manifests that houses’ masonry parts were not specific to the Phanariots but were modifications of a shared early modern Ottoman building type (taş odalar, i.e., masonry rooms). Near a special section of the prominent state bureaucrats’ houses, these structures could have been built as annexes to religious complexes, like schools or libraries. Similarly, in the modest Christian Orthodox and Jewish religious complexes in Phanar and Balat, the residential or educational annexes, as well as treasuries, were built with the same building type. The novelty in the early modern Phanar waterfront was the use of this type on the neighborhood scale, for residential use, and on the waterfront. 30 Furthermore, the geology of the extramural infill land enforced some site-specific planning conditions. The masonry wings were built on special expansive wood pile foundations and were limited to a certain point of the land infill, majorly on the roadside.
It has been articulated elsewhere that a typical Phanar waterfront house would possess two wings: a multi-story masonry wing toward the street and a timber-framed wing on the water-side. 31 The plots were defined by high side walls from the masonry wings to the sea. There was a courtyard combining the two wings and sometimes a garden, a landing stage, or a boathouse on the shore. The seaside was unassuming and painted in dark colors according to the Ottoman social code assigned to Christian properties. 32 Sometimes, the property could possess the small plots across the street attached to the fortifications used for services and stables. 33 There were some masonry buildings that could have been built attached to the inner or outer side of the fortifications; these, unlike the waterfront plots, grew horizontal to the street.
One peculiarity of the masonry neighborhood was that it, probably with the intentions of its owners, was kept cautiously out of sight. Until the mid-nineteenth century, there are hardly any images of the Phanariot neighborhood apart from some panoramic views drawn from afar. In 1851, however, the same year of Byzantios’ publication, the Ottoman Ministry of Defense would acquire a new map of the walled city Istanbul, known after its engraver Olivier, which would use a special denotation for the Phanar masonry neighborhood 34 (Figure 1). The waterfront in front of the Golden Horn fortifications was shown as a continuous, thin, enclosed urban block with public access only at the landing stages across the city gates (six gates from Cibali to Balat, including Aya, New Aya, Petri, and Phanar). Moreover, the map had a different specification for Phanar waterfront houses’ masonry wings. In certain sections, several masonry wings were shown as a continuous street façade, whereas in other places, they were shown by singular square shapes. 35 Why these masonry buildings were shown on this map in the first place deserves to be an inquiry by itself. It is also a question whether there were more masonry wings before 1821 since some could have disappeared in the thirty years after the Phanariot’s houses’ confiscations. Whatever the reason for signification and how many Phanariot masonry wings had survived, the 1851 map remains, as far as this research is concerned, the oldest visual primary source to document the masonry neighborhood as Gautier and Byzantios had witnessed in situ. Moreover, since what is observed in the 1851 map shows older buildings, some of which date to the late seventeenth century, it represents partial clues on the earlier urban fabric: the high street of the Phanariots.
Traces of Phanariots’ masonry houses on the 1851 Olivier map, detail.
On the precedent of the 1851 Olivier map, some later Ottoman state and municipal maps would use a similar specification for the Phanar houses’ masonry sections until the cadastral maps were produced in the early twentieth century. The continuity of specification provides a unique opportunity for making diachronic connections. These maps belong to a time when Phanar waterfront, as it remained and transformed, became visible in photographs and other visual documents.
Late-Nineteenth Century Interventions
After the mid-nineteenth century and through the Late Ottoman period, four major types of interventions transformed the larger Phanar waterfront from Cibali to Balat: post-fire partial plans, destruction of city fortifications, industrialization, and street enlargement. Ironically, these destructive interventions became drives for the documentation of the masonry neighborhood and provided the best data set for its visualization.
1863 Stolpe map, dedicated to Sultan Abdülaziz, depicted no major urban change for Phanar after the mid-nineteenth century. Here, like the Olivier map, the masonry wings of the waterfront houses were specified, with persisting numbers and distribution. Stolpe map denoted the confessional community distribution by color: Muslims in pink, Christians in gray, and Jews in yellow. In Phanar, properties still majorly belonged to Christian citizens 36 (Figure 2).
Traces of Phanariots’ masonry houses on the 1863 Stolpe map, detail.
The 1875 to 1882 so-called Ayverdi maps demonstrate that specific intramural sectors of Phanar and Balat had been devastated by fires where partial street regularization plans were realized in an orthogonal and grid-iron block pattern. 37 Masonry multi-story houses were built as attached properties on these new urban blocks (Figure 3). Being fire-proof, the Phanar waterfront houses’ masonry sections—also demarcated in these maps—were not majorly impacted by the fires. However, their timber wings might have been damaged, sometimes rebuilt, or left vacant. As intramural Phanar redeveloped into a dense Greek Orthodox neighborhood, the landing squares for the new steamboats expanded further onto the sea with small shops, coffeehouses, theaters, and taverns, as well as Police stations, which might have transformed some of the old masonry houses close to these locations. 38
Traces of Phanariots’ masonry houses on the 1875 to 1882 Ayverdi maps.
Ancient city walls’ destruction was an ambitious project of the Reform Period governments, which was first discussed in the 1850s and was conducted as a very problematic process with auctions, partial demolitions, interruptions, several documentation and mapping attempts, and, after half a century, concluded at most places without a change in the actual urban plot pattern and points of access. 39 In the Phanar region, apart from some flat sections around Cibali-Ayakapı and near Balat, the ancient fortifications were built on a sloping terrain, and being retaining walls, they were not demolished. In some parts, new masonry and wooden buildings were constructed attached to or replacing the wall; in others, the fortification was kept bare. 40 The preservation of the fortification line, whether by physical remains or by the urban plot pattern, meant the waterfront fabric’s isolation from the intramural sectors would continue but for another outcome. As the intramural Phanar flourished, the extramural waterfront would decay.
Phanar was at the inner section of Istanbul’s natural harbor, Golden Horn, across the Great Arsenal, one of Istanbul’s great industrial sites. 41 For building new factories and ateliers, the waterfront properties began to be sold to private developers. Some properties were left vacant for investment. Due to the demands of warehouses and workshops, the landfills increased. The Unkapanı-Cibali area was the first affected place as being closest to the actual harbor quays. The Patriarchate protested this process of industrialization since it damaged the communal character of the neighborhood, its spirituality, and its relationship with the Golden Horn. 42
The industrial facilities on the waterfront, as well as the redeveloped and densified intramural neighborhoods, brought a new strain to the street traffic. The old Phanar high street (then known as Phanar Avenue or Grand Phanar Avenue) coped to sustain horse coaches and carts’ movement. The project to enlarge the street, with cobblestones and sidewalks, by demolishing the properties on two sides came into the agenda. Demolitions were not realized at once, but the line of the enlarged street was pre-determined, and when a new building was to be constructed, they obeyed the new plot-street line. This factor left the standing masonry houses or plots with a historical past projected toward the street. Their once attached side façades exposed, they emerged as ancient relics. The best source to visualize this significant street profile as evidence is the 1913 to 1914 cadastral maps known as the Alman Mavileri. 43 These maps formed with geographical instruments, mark points, altitude, and triangulations constitute the earliest base maps that might be calibrated with present-day digital documents. Furthermore, they provide evidence of some Phanar waterfront plots, which had not been documented otherwise (Figure 4).
Traces of Phanariots’ masonry houses in the street line, 1913 to 1914 German Blueprints, cadastral maps known as Alman Mavileri, detail.
Awareness of a Disappearing Past in the Twentieth Century
The loss of built heritage caused by these urban interventions and the search for a historical urban identity brought Phanar to the attention of Western antiquarians, Ottomans, and specifically Istanbulite Greeks starting in the 1870s. 44 The main objects of interest were Roman and Byzantine archeology and building heritage. The Phanar waterfront houses attracted attention; their differentiation from the other historical residential buildings was recognized, and attempts were made to categorize them within existing art and architectural history cannons. There are several books on French and German debating on the Byzantine or Ottoman character of the houses, including visuals and drawings. There were also some books written by Greek scholars, such as Manuel Gedeon, clarifying the houses’ original Phanariot context. 45
Léon de Beylié’s “L'habitation Byzantine”s supplement volume “Les Anciennes Maisons de Constantinople,” published in 1903, is a primary source hypothesizing a Byzantine origin (both as provenance and stylistic influence) for the masonry wings of the Phanar waterfront houses. 46 His speculative art historiographical proposal aside, which is still influential to some extent in the present day, de Beylié’s work is a very important visual source, including several house plans and the earliest photographic portfolio of Phanar masonry neighborhood taken in a sequence from the main street 47 (Figure 5).
Plates showing Phanariot houses photographs from Léon de Beylié’s L’habitation Byzantine’s supplement volume “Les Anciennes Maisons de Constantinople,” 1903.
The second important source from the early twentieth century is German art historian Cornelius Gurlitt’s Die Baukunst von Konstantinopels, published in 1913, which included several plan, section, and elevation drawings as well as photographs from the street and interiors. 48 (Figure 6) Gurlitt described the wooden wings of the houses but did not provide any image from those parties. He was also the first to propose a stylistic and materialistic relation between these masonry wings and the seventeenth- and eighteenth-century Ottoman masonry buildings. Both de Beylié and Gurlitt were introduced to the Phanar houses by a local resident, Officer Romulus Spatharis, who, having two adjacent houses between Ayakapı and New Ayakapı (number 5 and 6), aided them in visiting the surviving houses and especially their interiors. 49
Pages and plates depicting Phanariot houses from Gurlitt’s Die Baukunst von Konstantinopels.
Several other late nineteenth- and early twentieth-century photography albums are important visual documents, such as panoramic views from the Galata Tower, the Pera cemetery and park, Petits Champs, and the Golden Horn. 50 These photographs taken from the same spots form a basis for comparative analysis, specifically to comprehend the timber-framed wings of the houses. A panoramic photograph taken by Officer Ali Rıza Bey in 1907 from the Sultan Selim Mosque platform above Phanar, in Sultan Abdülhamid Albums, is a very important visual source to represent the condition of the Phanar waterfront houses from the Phanar landing stage to Ayakapı (Figure 7). It demonstrates the continuing density of the urban fabric in successive reconstructions. The photographs of Eugene Dalleggios and Achille Samandji, from the early twentieth century to the 1920s, are forming a collection in Greece, which provides additional information for some of the preserved buildings. 51 The photographs of artist Avni Lifij, as published by Sedad Hakkı Eldem, are another source to show the houses in the 1920s before their demolition. 52
Panorama of the Golden Horn from Selim I Mosque.
In the 1920s and 1930s, after the foundation of the Republic of Turkey, a larger part of the Phanar waterfront was transformed into small industry and shipping functions. Final street enlargement works were realized with the partial demolition of some masonry houses. For this period, the Pervititich Insurance Maps are the major source and provide specific building information: construction materials, structure, and number of stories 53 (Figure 8). Industrialization would further increase after the 1940s when the French planner Henry Prost’s Master Plan defined Golden Horn’s inner harbor as a small industry zone. 54 Until the 1970s, the waterfront was totally transformed into industrial facilities, when the remaining Phanar houses were incorporated into these facilities or just left as a façade. Since most of the remaining Greek citizens left in successive stages after the September 6 to 7, 1955 Istanbul pogrom, the property rights issues might have resulted from abandoned and unoccupied buildings, the traces of which are preserved in property patterns.
Phanariots’ masonry houses, detail from Pervititich Insurance Maps, 1929.
Published in 1980, Sedad Hakkı Eldem’s two-volume book on Ottoman residential architecture titled Turkish House, which includes his earlier in situ documentations of the Phanar masonry neighborhood, interpreted these buildings as a variation of Ottoman architecture, the so-called masonry rooms (taş odalar) (Figure 9). 55 The Golden Horn harbor revitalization projects of the 1970s and 80s brought a new interest in Phanar. Teams from the Academy of Fine Arts (Mimar Sinan University after 1982) initiated by Professor Haluk Sezgin documented the Phanar shore before the waterfront park projects of Major Dalan in the 1980s, which cleared all the industrial facilities, including several of the remaining houses. 56
3D modeling of a Phanariot house by the authors after Sedad Hakkı Eldem’s drawings.
The municipality’s digital maps were drawn in the early 1990s, and they show the Unkapanı-Balat waterfront after the Major Dalan projects and the construction of the waterfront parkways. These maps have dotted line marking properties, which is useful in adjusting with the previous maps. 57 With the start of Fener Balat’s urban renovation projects by the late 1990s, there was a new focus on the area, but since the waterfront fabric was almost destroyed and replaced by parks, it was intramural sectors of Fener and Balat that have been worked on. 58
Phanariot Heritage in Situ
What traces have survived this long disappearance history? Eleven buildings are standing from the Phanar masonry neighborhood 59 (Figure 10). Five of these have been recently restored by the Istanbul Greater Municipality and opened as art exhibition venues. 60 Another has been used as the Women’s Library since the 1990s. 61 Three surviving religious institutions have their annexes in the Phanar masonry building typology. The Saint Nicholas Church near Ayakapı on the outer fortification side has a masonry hall over a holy spring or hagiasma (Agios Charalampos). 62 The Bulgarian Church has a large service building constructed in the 1840s attached to the fortifications and one tower. 63 The Sinai Monastery metochion (ecclesiastical embassy church) possesses two masonry wings, one of which is the most well-known Phanar masonry residential unit since it was pictured by many Western travelers in the nineteenth century. 64
Houses standing in situ documented through site surveys.
Since the remains on the waterfront are limited, similar masonry buildings in the intramural sections provide comparative evidence -neighborhoods from Cibali to Balat around Seferikoz, Gül (former Saint Theodosia), Sultan Selim and Fethiye (former Theotokos Pammakaristos) mosques, Boğdansaray (Moldavian Palace), Balat market, Phanar and the Greek Orthodox Patriarchate. Fourteen such masonry buildings can be identified, most important of which are preserved as part of religious complexes while few are within the commercial streets. The Greek Orthodox Patriarchate houses a three-story masonry tower structure, probably from the seventeenth century, used continuously as a treasury, and a two-story horizontal masonry building attached to the tower. 65 The Panagia Muhliotissa Church (Kanlı Kilise) has masonry annexes in the courtyard. The large plot, which houses the Saint George Church (Jerusalem Holy Apostles’ representative), between Vodina and Sancaktar streets, is significant, as the so-called palace of Dimitrie Cantemir is located here. 66 The residence was constructed in the seventeenth century and is formed of a lower two-story masonry building and a single-story horizontal masonry mass constructed above a high, pointed arcade reached by an open monumental staircase. In the neighboring Balat quarter, two synagogues (Ahrida and Yanbol) and an Armenian church (Surp Hireşdagabet) possess masonry annexes. 67 The intramural masonry structures are seventeenth-century precedents for the masonry wings on the waterfront.
Mapping and Visualizing the Phanar Waterfront
Diachronic approaches aim to reveal the spatiotemporal changes in the urban form by layering urban fabrics over time. 68 By juxtaposing layers from different periods, a simulation of “urban sedimentation” is produced to decipher models of growth, evolution, or transformation in the history of urban form. 69 The conceptualization of the city as a repository of successive transformations is elaborated by French urban sociologist Henri Lefebvre as follows: “Each urban formation knew an ascent, an apogee, a decline. Its fragments and debris were later used for/in other formations . . . Destructurations and restructurations are followed in time and space always translated on the ground, inscribed on the practicomaterial, written in the urban text, but coming from elsewhere: from history and becoming.” 70 Following Lefebvre’s definition of the city as a “projection of society on the ground,” urban form can be traced, analyzed, and reconstructed by its inscriptions on historical maps. For the diachronic analysis of the Phanar waterfront, maps of Istanbul dating from 1851 to 1928 were the primary visual sources.
Cross-referencing data addressing the same period or building in different sources and evaluating the consistency of data from different periods were significant in managing different levels of spatial granularity in the process of spatializing historical evidence (Table 1). The complexity of the spatialization in the data application process was elaborated in three levels—diachronic analysis, base map production, and 3D modeling—that were methodologically reciprocal. Data application studies were realized simultaneously at two different scales, building and neighborhood. While a digital collection of house models was developed based on primary sources, the 3D models of houses were simultaneously installed in the digital site model of the Phanar neighborhood.
Primary Visual Sources and Their Information Coverage on the Phanar Waterfront and Houses, shades correspond to the availability of information.
A Situated Inventory: Diachronic Analysis of Cartographic Evidence
The data potential of late nineteenth-century maps was mentioned above, which are not perfectly scaled drawings but are important to represent major urban elements and the continuity of the masonry houses. It was possible to achieve a level of consistency with the later reference-marked maps, among which building footprints and street outlines could be followed and compared by approximate juxtapositions. The Alman Mavileri cadastral maps of 1913 to 1914, the earliest reference-marked sources, were selected as the ideal historical source to produce a digital base map onto which the data acquired from both earlier and later periods can be transferred and installed. The hypothetical base map was digitally produced by the polygonization of the Alman Mavileri, and the process started with tracing the constants, or rather the historically persistent elements, of the urban fabric as these elements strongly determine the organization and transformation of the urban form 71 (Figure 11). Tracing the border defined by the Golden Horn Sea Walls, with the gates and towers marked along the city walls, was fundamental, as well as the historical monuments such as Gül Cami (former Hagia Theodosia), Sultan Selim Mosque, the Greek Orthodox Patriarchate, and so on. The monumental structures were also mapped according to their existence in different periods. 72
Phanariots’ masonry houses traced by superimposing the 1913 to 1914 Alman Mavileri and the 1928 Pervititch maps. Map fragments assembled, scaled, superimposed, and tracings added by the authors.
Another significant border to be traced was the shoreline, yet it was not one of the constants but rather the transforming element in different layers of diachronic analysis. 73 By juxtaposing multiple shorelines, it was possible to document the amount of infill change from the 1850s to the 1930s (Figure 12a). Second, the street network and residential urban fabric, which remained mostly unchanged along the fortifications through the layers of diachronic analysis, were drawn. Third, the Phanar houses, or rather their masonry parts, which were discovered and documented through data collection, site survey, and diachronic analysis, were drawn on the base map. Finally, possible property lines were traced based on the marks and similar plot layouts observed in the diachronic layers of the urban fabric as a hypothetical layer to illustrate the ownership pattern of the extramural waterfront (Figure 12b).
(a) Diachronic analysis of changes in the shoreline. (b) Phanar digital base map showing (i) Golden Horn fortifications, (ii) urban blocks and streets, (iii) shoreline, (iv) monumental buildings and religious complexes; (v) Phanar waterfront plots; and (vi) Phanar houses. Produced by the authors.
In the larger area from Cibali to Balat, each early modern masonry residential building in primary visual sources, observed during site surveys, discovered through the diachronic analysis, and finally installed onto the digital base map is labeled and cataloged according to their location, typological formation, and function. 74 The initial inventory of the Phanar houses started with eleven houses, reaching a total of fifty-five at the end of the data application process. With other masonry buildings in intramural neighborhoods, the number of annotated units reached seventy-four. The numbers and distribution of the houses are presented in Table 2.
Numeric Inventory of the Phanar Houses.
Diachronic analysis and the derived base map communicate information in multiple levels and scales of architectural type and urban form. Reading “architecture as evidence” 75 and buildings as cultural artifacts, the digital base map does not simply provide a 2D layout for 3D modeling but rather constructs a situated inventory, or a database, of the Phanar houses in the mid and late nineteenth century. On the other hand, the digital base map is not a passive repository of data; it is constructed and constantly reconstructed by reasoning and corroboration as it examines and records clues and traces in the built fabric.
Studies in urban morphology assume that there is “a particular logic” or “a systemic organization” that has determined the organization of the urban fabric by constant or permanent elements and rules of transformation, causing changes in the fabric in different periods. 76 Therefore, the development of the urban fabric is never random, but there are morphologically identifiable rules or laws that can be extracted from it. Such a hypothesis, at the core of morphological analysis, is built upon the interdependence between building type and urban fabric. 77 The urban morphology of the Phanar extramural waterfront and the building typology of waterfront houses display such interdependence from the seventeenth to the late nineteenth century. The interdependence between part and whole, in other words, between typology and morphology, makes Phanar waterfront and the Phanariot houses a unique case to illustrate the specificity of a building type and its resonances in urban plots that have been morphologically and historically compacted between the city walls and the sea. The masonry residential units were also constructed in the intramural Phanar and Balat, mostly as part of religious complexes, by the inner sections of the fortifications, and by shopping streets. However, these units differ in their urban plot patterns from the waterfront. In this regard, the digital base map stands as a historical hypothesis on the urban morphology of the Phanar extramural waterfront that is open for testing and improvement.
Reconstructing the Phanariot Houses: Visualization by Architectural Modeling
Simultaneous with the diachronic analysis and base map studies, a collection of 3D digital architectural models of houses on the waterfront constituting the masonry neighborhood, was produced. It was possible to collect enough data to digitally reconstruct thirty of the fifty-five houses, yet the level of historical evidence for 3D modeling and degree of reliability differentiated for each house.
Historical evidence varied from primary visual sources such as architectural drawings, photographs, and historic maps from different periods to primary textual sources such as first-hand descriptions and surveys of the houses. While still-standing houses could be studied in even more detail by in situ surveys and their recently produced restoration drawings could be acquired, a photograph showing a single façade, and a trace detected on a historic map had to be interpreted for some of the demolished houses. Accordingly, five degrees of reliability were defined regarding the accuracy and completeness of 3D models of the houses based on the level of historical evidence collected (Tables 3 and 4; Figure 13).
Degrees of Reliability Determined for Modeling the Phanar Houses According to the Accuracy and Detail of Historical Evidence Collected from Primary Visual Sources and Site Surveys.
Degrees of Reliability of the Phanar Houses on the Extramural Waterfront.
E: exact, A: approximate.
Houses on the Phanar extramural waterfront according to degrees of reliability.
These varying degrees of reliability illustrate that each house’s modeling process was unique since each 3D model was produced in reference to and by correlating historical evidence in different levels of spatial and visual granularity in the same scale. The modeling process of the houses in the first degree of reliability was straightforward. Plan, section, and elevation drawings of the houses are digitally re-drawn in two dimensions, and then these drawings are re-positioned in three dimensions in reference to each other. 3D models of the houses were produced by extruding plans, articulating interior configuration according to sections, and detailing the openings and roof formation according to façade and section drawings. The modeling process of the houses in the second degree of reliability followed a similar method by producing 2D architectural drawings first; however, the dimensioning was not as precise as in the first degree. Although architectural drawings published in primary sources usually had scale bars, the overall dimensions had to be cross-checked by proportioning plans to the traces on the historic maps or by comparing the height of the houses in the façades and sections considering perspectival distortions in photographs (Figure 14). The modeling process in the third degree proceeded directly and solely in three dimensions. As it was not possible to produce precise architectural drawings in plan or section, an approximate footprint of the house is determined according to the traces detected on historical maps. Then, the overall mass of the house is extruded by proportioning the façade of the house in reference to adjacent houses that are previously modeled and seen in the same street photograph. Similarly, floor heights and openings are proportionally determined on the façade, and details are added according to the architectural rules and elements extracted from previously modeled houses. Houses in the fourth degree of reliability are modeled as abstract masses without any façade or interior details due to the scarcity of visual data. Finally, houses with a fifth degree of reliability could only be shown as translucent ghost-like masses as these houses’ traces could be partially detected on historical maps.
Instances from the modeling process of houses with first and second degrees of reliability.
When all the extramural waterfront houses modeled are brought together, a typology of the Phanar masonry neighborhood becomes visible (Figure 15). Façades display strong proportional correlations, and it is possible to extract a catalog of architectural elements, details, and materials such as alternating stone and brick courses, arched gateways, windows, stone windows frames with pointed or rounded brick arches on top, projections of upper floors with stone corbels, pitched roof with brick saw-tooth cornice (kirpi saçak), and so on. 78 The minimum façade width is 5.2 m, and the maximum is 19.2 m. The number of stories is between two and three, and the height changes between 9.3 and 14.5 m. It has been noticed that some buildings with larger façades were formed of two interlocking houses. Courtyards on the street-side are rare, but when they exist, they occupy the width of a house plot. The projecting parts can include the whole building, but when some part is extruded singularly, the projection widths change between 3.7 and 5.7 m. A comparison of sections also reveals details about the unexpected interiors behind the façades, the most glorious of which are the well-hidden domes or cross-vaulted domes that characterize the architecture of the upper floors (Figure 16). Further architectural and historical readings and stylistic analysis of the houses can be produced beyond the scope of urban studies of this paper. 79
Street façades of the Phanariot houses—A typological comparison.
Section perspectives of the Phanariot houses.
Visualization on the Neighborhood Scale
Following the production of the digital base map of Phanar and the house models toward the 3D urban model, the composition required the additional data layer of topography to combine two and three dimensions. The topographic information was acquired from the layers of diachronic analysis (1913-1914 Alman Mavileri cadastral maps, 1928 Pervititich insurance maps, and 2018 Municipality maps; Figure 17; Table 5).
Key map for spot elevations along Phanar Street.
Topographic Changes Along the Phanar Street (Fener Caddesi) on the Extramural Waterfront in 1928 and 2018.
The topographical data on the Phanar high street is a specific concern since the street levels have changed considerably here. This was a long-enclosed artery, where the surface water from the intramural sectors flowed with few accesses to the sea at the landing stages. Since the street was narrow and facing north, it was damp, and the collected mud created a constant rise on the street level. Table 5 shows the changes in the elevation of Phanar main street by comparing the spot elevations given on maps from the 1920s to the present. It is possible to distinguish three intervals along the street in reference to the rate of level rise. While on Spot 1 and the area defined in between Spots 6 and 11, the rise of street levels changes between 0.3 and 0.7 m, the street level rises to 1.7 m on the interval between Spots 2 and 5. These intervals, defined by the street level rise rate and the urban fragments identified by the distribution of existing and demolished houses, match each other. In other words, wherever more houses were demolished, the rate of the rise in street level was higher. To illustrate, there are no existing houses on the urban fragment between Aya and Phanar gates except House E2, which still stands in situ. 80 This extreme rise in the street level can be observed in situ, where the ground floor of House E2 is almost below the street level today. On the other hand, the urban fragment between Phanar and Balat gates contains most of the existing houses where the street level rise decreases below half a meter.
The addition of the topographic data layer to the digital base map was a major step in modeling the Phanar neighborhood. A patch surface was generated according to the 3D coordinates of spot elevations. Then, the polygons of streets, city walls, the shoreline, and building footprints drawn on the base map were projected onto this topographic surface. As the final and the least reliable layer, other residential buildings are modeled to achieve a tentative representation of the intramural urban fabric (Figure 18). It should be noted that this layer is not part of the Phanar extramural waterfront and is highly speculative. Even though most of the street network and urban blocks that are continuous from 1875 to 1928 can be determined through the diachronic analysis, the built fabric cannot be precisely known since the earliest source noting the number of stories is from the 1920s, the Pervititich maps. In the current 3D model, the built fabric was modeled by an algorithm that extrudes the polygons outlining the building footprint up to the building heights annotated as spot elevations on the 2018 map. The model develops a conceptual visuality to balance the nuances coming from historical evidence in varying degrees of certainty and reliability. In this regard, the rendering of the 3D model visualizes the context as yet-to-be-fixed and malleable.
Views from the 3D digital site model.
To further explore the potential of the 3D model, or better to say the potential of a 4D model, as a research tool, samples of partial and large-scale views were produced. Partial views concentrated on comparing streetscapes with the original photographs to better understand the distinctive relationship between urban form and building type (Figure 19). The street views are a check for reliability and a possible canvas for future research to further complete the missing urban fabric with additional data and/or graphic abstraction.
Comparison of 3D model views with various original street photographs.
On the other hand, larger-scale views were taken from the points that can be compared with the nineteenth- and early twentieth-century visual materials. Panoramic view from the geographical location of Petits Champs, a significant historical observation point, has been produced from the 3D model and juxtaposed with the nineteenth- and early twentieth-century photographs taken from the same point. These juxtaposed views show the Golden Horn façade within which certain houses can be annotated (Figure 20). The masonry neighborhood was a timber neighborhood when it was witnessed from the harbor and the paper aimed to restitute the masonry neighborhood rather than the waterfront. However, for future research, specifically at the sections closer to the observation point, the Cibali-Petrion waterfront, the houses’ timber-structured façades can be annotated with some precision. These representations are forming a basis for future research to further differentiate the timber-framed wings of the Phanar waterfront houses by studying each house individually.
Comparison of 3D model view with the panorama from Petits Champs des Morts Pera.
Outcomes of Phanar Waterfront’s Visualization
The base map, the 3D site model of the Phanar neighborhood, and the 3D models of individual houses are the present outcomes of the research. The complete model of the Phanar neighborhood contains the following elements: (1) topography, (2) street network, (3) city walls, (4) monumental buildings, (5) Phanariot houses, and (6) a tentative inner-city built fabric. Yet such large-scale modeling of a historic urban context inevitably inherits an instability embedded in varying levels of historical evidence and spatial granularity. In other words, the degrees of reliability defined for the houses also reflect on the urban scale.
For the Phanar high street, it is possible to define five intervals matching with the city gates’ locations: (1) Cibali—Ayakapı; (2) Ayakapı—Yeni Ayakapı; (3) Yeni Ayakapı—Petrion; (4) Petrion—Phanar; and (5) Phanar—Balat. Starting from the West, the interval between the gates of Balat and Phanar can be assessed as the best-known urban fragment due to the higher number of existing houses and houses with a second degree of reliability. The existence of properties belonging to the Sinai Monastery metochion and the Bulgarian Exarchate has prevented interventions on the waterfront, and the adjacent properties were preserved altogether in successive urban interventions. The fortifications are also preserved partially, with some fortification towers providing the original street scale. However, it should be stated that this section is the longest part of Phanar Street, and except for the Sinai Monastery area, the masonry houses display a fragmented street pattern. It is a possibility that some properties had already been demolished by the mid-nineteenth century, such as the case of the Vogorides mansion that was destructed to be replaced by the Bulgarian Exarchate. The interval between Phanar and Petrion is the most conjectural, where there are only houses with third and fourth degrees of reliability. Being in front of the Patriarchate and the most prestigious location, the lack of evidence is significant. Nevertheless, the distribution of documented properties in the plan shows a similar pattern to the Balat-Phanar section. Even though the urban fragment from Petrion to Cibali encompasses three intervals with varying granularity of historical evidence, they mostly contain houses with a second degree of reliability with rare instances of existing houses. Ayakapı and New Ayakapı areas can provide and are significant to display the most complete street restitutions. The question posed by the different degrees of preservation as it was recorded after the mid-nineteenth century is whether there were more masonry houses, such as in the New Ayakapı sector, which can only be given by a thorough study of the pre-mid nineteenth-century historical sources, such as confiscation registers. As a preliminary note, it can be stated that in Bostancıbaşı registers between 1821 and 1824, the confiscated properties are mentioned for the Yeni Ayakapı-Balat section and match with the places where the urban fabric is fragmentary and incomplete.
Apart from the waterfront, which was the primary focus, some information on the intramural sectors was also achieved. Although a singular 3D modeling of these structures was not aimed at this stage, they were included as masses within the large-scale 3D site model. Some urban patterns in these intramural areas have been noticed and can be mentioned here. The masonry building blocks dominant on the waterfront also continued on the inner side of the fortifications. On the inner face of the Petrion fortifications across the Patriarchate, some attached masonry houses were located. Similarly, on the inner face of the fortification by the Phanar and the Diplophanarion gates along the market street, there were some attached masonry buildings with shops on the ground floor. On the street near Saint George (Holy Sepulcher metochion), today’s Vodina Street, there seem to have been other masonry houses with shops below. Third, the annexes of Yanbol Synagogue and the Armenian Church define a street formed of masonry buildings.
What does the urban typology of the masonry neighborhood suggest? That is a mile-long extramural waterfront fabric of attached long and thin residential plots that are only interrupted at six points of public access and are made of half-masonry and half-timber houses. Attached narrow and high buildings forming continuous plots is an urban type often seen in Mediterranean historical port cities, at least since the Middle Ages. 81 Some cities within the Phanariot network, in the Aegean and the Eastern Mediterranean, had possessed similar waterfront buildings, urban plots, and/or masonry structures. 82 Chios, the original hometown of the prominent Phanariot family Mavrokordatos, for example, had a considerable residential masonry architecture, both at the harbor and the suburban Kampos. 83 The cities of the Levant, like Haifa (Jaffa), and the port of Jerusalem had ancient masonry building traditions. 84 The residential architecture of Rosetta (Rashid), Damietta (Dimyat), and Alexandria constituted of multi-story brick masonry buildings. 85 The value of land, the commercial potential, and the separation of the harbor fabric from the inner city by fortifications or plot patterns are determinants for this kind of morphological formation. In Istanbul, there were commercial neighborhoods on the Golden Horn that had similar plot types, such as the Ayazma Gate Quays. However, Phanar was not in the commercial part of the Golden Horn and should be interpreted separately from urban plots in the working harbor. The Phanariot houses belonged to the mercantile elite, but the commercial potential of these houses could at most be connected to storing some merchandise on the ground floor, maybe comparable to the ports mentioned above or Venetian mercantile canal houses. 86 However, the textual evidence for this hypothesis is still incomplete and scarce.
The urban plot type, made of very thin plots extending onto the harbor along a main artery, has a comparative case in İzmir, the second important port city of the Ottoman Empire, Frank Street. Frank Street was made of buildings as warehouses and residential units (houses of Franks, frenkhanes or verhanes) along a main street, each working as a separate landing stage, a comb-like urban pattern. 87 The commercial character of the İzmir case poses a difference from Phanar. Yet, the compactness, attached property type, and linear expansion of plots by land infill show urban morphological similarity. Maybe the best way to compare the two atypical streets in two Ottoman cities is to define them as waterfront streetscapes with unique building types belonging to specific communities, that is, Frank merchants and Phanariots.
The distribution of the masonry wings of Phanar houses in the visualization project depicts that in the time of Gautier and Byzantios, the Phanar masonry neighborhood, that is, the Phanar high street, was not solely made of masonry houses, that is, kargirs. In some sections, there were a series of adjacent masonry buildings; in others, there was an even distribution of units to suggest a dominance of this type in the neighborhood. Furthermore, for the timber-frame houses, in between or the front of the masonry ones, the ground floors were also made of masonry, or there were masonry court walls, which should have increased the impression of a neighborhood made of stone. The historical myth of the Phanariots could have filled the blanks in the masonry streetscape—once prosperous and mighty, then fallen and in ruins—and the existing buildings could have been perceived as the remains of a more complete stone picture, such as in Ayakapı and New Ayakapı section. The suggestion that there could have been more masonry houses before the mid-nineteenth century has already been provided. The street façade was made of the same type of buildings, but each of these buildings was unique in height, width, façade articulation, and style. As suggested by Gautier, some of these variations might be related to the construction dates of the buildings. The blank and two-storied masses are usually from an earlier date; the ones with curved façades point to the eighteenth century and its baroque modes. The variations might also point to the ambitions of their owners, singular innovations they sponsored, and their status within the Phanariot society, in a similar pattern to early modern mercantile and bourgeoise communities. These patterns from the neighborhood scale can only be ascertained by further studies of each documented house individually. The line of Phanar Street followed the soft curves of the bay and the fortifications parallel to it, which provided constantly changing perspectives. The dynamic streetscape should have increased the dramatic effects of each masonry house’s façade, a phenomenon that Théophile Gautier perceived as faces, with teeth and lips, licking the masonry neighborhood. The stone neighborhood was Phanariots personified in brick and stone, and the present paper attempted to imagine it.
Concluding Remarks on Research by Visualization
Criticizing historians’ journey along the stream of truth, Burke suggests “to replace the idea of sources with that of ‘traces’ of the past in the present.” 88 He underlines the impossibility of a pure account of the past and argues for using images not solely as “evidence” but also for their power to inspire “historical imagination.” 89 This paper implemented a source-based methodology to investigate the disappeared urban fabric of Phanar’s waterfront by dwelling on the potential of visualization to enhance historical imagination. It instrumentalized the 3D model as a research tool onto which diverse traces of the past, that is, historical evidence, can be installed and integrated to foster processes of knowledge production. Yet, the evidence-based process of historical visualization is not a systematical act of fabricating a single accurate image of the past but a heuristic process of balancing interpretation and evidence. As multiple traces contaminate each other, the inconsistencies and scarcities in historical evidence require a scholarly method to refrain from overinterpretation while maintaining the traceability of documentation and visualization. This paper demonstrated the potentials, challenges, and limitations of 3D modeling in urban history research by configuring an epistemic setting with methodological and critical tools for documentation, analysis, visualization, and assessment, such as diachronic analysis, situated inventory, degree of reliability, granularity, and reality effect.
Recognizing the complex disciplinary field in which research by visualization can be anchored, this research contributes to establishing source-based digital reconstruction as a scholarly method in urban history and digital humanities. Yet, the methodology is never self-governed; it is constantly informed and improved by the contextual originality of Phanar’s extramural waterfront. The tenacity between the building typology and urban morphology of the waterfront reflects on the methodology to introduce tools for tracing the disappearance of the historical urban fabric, which will, in turn, serve to reverse it. As the resonance between the case and the methods facilitates confrontations between the past and the present, historical research processes expand into a 4D field. While the historical evidence and the 3D model interchangeably act as sources and tools for historical investigation, the narrative of disappearance becomes a foundation for visualizing the masonry neighborhood of Phanariots, the missing fragments of which are yet to be assembled.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article presents the outcomes of the research project titled Reinstalling the Phanariot Houses in Istanbul: An Architectural and Urban History Research by Visualization, conducted by the authors at TED University from September 2022 to September 2023. The project (T-21-B2010-90082) was supported by the TED University Institutional Research Fund (TEDU-IRF).
