Tracing the lost streets: a spatial mapping of Tokyo at the turn of the eighteenth century

The appearance of downtown Tokyo, particularly Nihonbashi—the bustling financial district comprising high-rise building complexes, department stores, and business setups—conveys the modern face of Japan as well as its architectural and urban characteristics. Before the city of Edo became Tokyo, the city streets' image had changed its physical identity at least twice: first, in the wake of the Meiji reforms of 1868 and then after the Great Kanto Earthquake of 1923 (Jinnai, 1995). The Great Earthquake’s aftermath gave way to the dramatic transformation of building culture, modernist urban planning, and Western influences, utterly wiping away the urban character that existed under the Tokugawa Shogunate (1603−1868).

The sense of place on the streets of Nihonbashi is now a far cry from how the district looked at the turn of the eighteenth century (Figure 1). In the earlier era, Nihonbashi was the place of the lower social strata forming the commoners' district: a square grid plan shaping the urban blocks, two axial arteries, wooden townhouses lined up along the urban front, wholesale outlets and workshops of all kinds, and, on average, two to three alleyways or narrow pedestrian passages that provided access to the backstreet housing forming the urban blocks' inlands (Sorensen, 2002).

Although the vanishing spatial culture of Edo has made research challenging, it has inspired significant research interest among both Japanese and Western scholars. Traditionally, studies have inspected fragments of material culture, archaeological findings, and archival resources and have also examined in detail urban development (Ichikawa, 1994; Sorensen, 2002; Hein, 2010; Jinnai, 2020); normative and legal constraints shaping building techniques and traditions (Coaldrake, 1996; Hatano, 1998); architectural evolution (Itoh, 1972; Engel, 1985; Morse, 1961); the relationships between economic and demographic booms and space use, and urban segregation (Takashi, 2000), among other aspects.

Most previous studies have focused on the built environment, which is the material basis for everyday life. A neglected area in the field is the spatial analysis of dialogues between the built environment and human agency; understanding the character of city streets is incomplete unless it includes perspectives on everyday life and embodied experiences. Indeed, unrevealing spatial practice is not only about the physical construct but also the lived and perceived experience. The interlocking of different layers of meaning, as in the French thinker Lefebvre’s (1991) spatial triad theory, is based on the integration of three legs: the spaces are perceived, lived and represented.

Lefebvre’s theory is especially interesting for historical spaces and can be an analytical tool for assessing them to understand what it is like and what it means (Briercliffe, 2015). In line with this, historic urban landscape approach (UNESCO, 2011) reflects urban areas in the widest context, including the historical layering of cultural values and attributes, covering the site’s land-use patterns, spatial organization, perception and visual relationships, as well as social and cultural practices, among many other features. It also encourages research that comprehensively extends the intangible dimensions of urban heritage by applying a range of innovative tools to document and understand their meaning for communities.

In particular, for lost urban spaces, as in the case of the entirely wiped-away traditional streets of Edo, it is almost impossible to provide an exact reproduction of the particular sense of the place. Nevertheless, reflecting on digital methodologies in architectural and urban history, this study attempts to bring the lost spatial culture and social city together. Based on the concept of “the city as one thing,” it bridges the physical and social city in tandem (Hillier and Vaughan, 2007). This study investigated spatial forms, human activities, interactions and movements within city streets. As city streets are directly linked to social life, they are the grounds for the daily rhythms of human agency and drive patterns of behavior through spatial use (van den Heuvel et al., 2020). In particular, the character of the building–street interface reveals the liminality of the urban space, along with the translation of prescriptive norms into built forms and spatial use. By creating boundaries and generating hotspots, the interface defines how people move and interact with each other.

To enable a close to complete and accurate hypothesis on how spaces formed interrelations and created spatial experiences within the built environment of the commoners' district of Nihonbashi in Edo (Tokyo) at the end of the eighteenth century, this study created digital simulations of long-lost urban spaces, making them visible. It attempts to reveal the unknowns about lost urban space, seeks a multifaceted evidence-based investigation by mapping spatial layers and analysis through space syntax, and demonstrates the versatile visualization of architectural and land-use regulations on urbanity.

The streetscapes of eighteenth-century Nihonbashi that are no longer extant represent the everyday life, culture, town planning, land use, and architectural culture of Edo. Located at the heart of Edo, Nihonbashi brought together social groups and political authorities. The district held a unique significance for urban space, everyday life and physical mobility in historical Edo. Intellectuals and creative people vitalized Nihonbashi’s economic life (Jinnai, 1997), and the neighborhood grew in the hands of its resident writers, artists and publishers as it transitioned economically and socially toward the experiences and tastes of the average common consumer (Yonemoto, 1999). Nihonbashi’s streets were alive with a constant flow of people, goods and social interactions. Status- and class-based social orders blurred amidst the crowd on Nihonbashi Street as clerks mingled with officials and beggars with warriors (Figure 2).

Daily activities affected the spaces' fluidity, and their spatial presence was in transition. In Nihonbashi, two-story townhouses frequently stood adjacent to each other, occupying juxtaposed plots facing the main street. The architectural culture of traditional Japanese housing and a strong sense of commonality in urban spaces, particularly along Nihonbashi Street, were dominant. Typically, although not uniformly distributed, townhouses followed a repetitive design pattern and appeared cohesive. For example, architectural components tend to follow a typical pattern, from shop sliding doors to the latticework of windows. In particular, the absence of any permanent physical attachments, such as hinges on the front façades, contributed to the flux and flexibility of street spaces (Bharne, 2014). Shopkeepers along the street opened sliding doors and curtains at the shopfront to extend the shop space into the street space, seamlessly blending it as an integral part of it. When closed at night, the space was restored to the street and regarded as part of it (Saygi and Yasunaga, 2021, quoted from Tamai, 1986). This characteristic shaped a building culture that was dramatically different from European street spaces by creating flexible boundaries.

Saygi and Yasunaga (2021) explored the spatial culture in Nihonbashi, focusing on a long-lost “high street” by hypothesizing in three-dimensional space and employing a source-based approach in its digital reconstruction. Prior research has scarcely commented upon the contrast between the commercial urban front and the densely packed alleyways of tenements in the back; as Gaubatz (2016) also suggests, there is little research addressing the interesting liminality, public/quasi-public/private nature of the spaces, or the divisions and the boundaries. Indeed, the internal areas of urban blocks housed most of Edo’s population. The row houses lined up along the back alleys behind the main streets and within the urban blocks were spatially discrete but densely populated. Row houses represented a typical way of life among the lower social tiers. Few researchers have focused on how inhabitants navigated city streets or on what they visually experienced along the commercial strip.

Today, authentic Japanese woodblock prints are relatively common and frequently reproduced. Many prints portray Japanese life in Edo and beyond, revealing urban life within the closed Eastern nation for over two centuries. Especially after the second half of the eighteenth century, with the Kansei Reforms, social and cultural policies gave rise to urban culture and literature. This also reflects the blossoming of various printed materials showing what Edo streets looked like before the Meiji Reforms. Woodblock prints from this era depict famous places narrated in literature, as well as urban spaces, and provide a witness to historical events, landmarks of the city, places of entertainment and seasonal activities such as cherry bloom viewing (Edo-Tokyo Museum, 1996). These prints also provide broad observations of daily life, urban spaces and city dwellers, revealing many aspects of the city’s spatiality that have now been lost.

Although these visual depictions are rich and significant resources, to what extent do they show how the city was visually experienced and navigated through human agency? As Jacobs (1961, p. 30) states, “Great cities are, by definition, full of strangers—even residents who live near each other are strangers and must be because of the sheer number of people in small geographical compass.” Were the commoners in Edo involved in each other’s lives because of the spatial togetherness along the back streets? Did spatial configurations allow residents to preserve privacy? How did Edo handle strangers? Gate regulations and mobility restrictions during the night are well documented. However, what about the daytime? To what extent did residents have “eyes upon the city streets” to watch street activity, provide surveillance, and create an “unconscious network of voluntary controls” (Jacobs, 1961, p. 32)? The commercial sphere of the main street is continuously used by humans. Although there is a clear distinction between the main street and the alleyways, how well were the two integrated for the inhabitants regarding visibility, accessibility, and mobility?

In the case of Edo streets, answering questions about everyday experiences and human agency can be cumbersome. The urban transformation process during the transition from Edo to Tokyo presents difficulties in tracing the historical material culture; the earlier ongoing demolition and reconstruction due to physical damage from numerous fires and earthquakes also make it challenging to capture (Sand, 2012). The fires not only resulted in swapping away the built environment but also caused a significant loss of archival documents, such as population registers (Hayakawa, 2014) and cadastral maps (Iwabuchi, 2017). However, some historical documents such as those preserved through Mitsui’s bookkeeping practices remain. Records from the Mitsui Bunko Archive, particularly an archival document showing the land use of a property owned by Mitsui and indicating building layouts, measurements and ownership, have been fundamental to this research to uncover the spatial features of a typical urban block in Nihonbashi, Edo.

This study examines a sample area: the neighborhood of Honshirogane-chō 2 chome, as it was named in the eighteenth century; today, this area is known as the Nihonbashi Muromachi 4-chome, Chuo-ku, Tokyo. The area covers a typical urban section at the intersection of two arteries in the commoners' neighborhood of Nihonbashi, approximately 500 meters north of the Nihonbashi Bridge, near the famous fish market. The sample area includes the crossroads of the prestigious high street that provided solid economic advantages for merchants—Nihonbashi Street within the north-south axis—and the street along the Kanda moat that connected the castle to the Kodenmachō prison and the execution ground—Honshirogane-chō Street within the east-west axis (Figure 3).

The discussion in this paper is based on archival data on the neighborhood, visual depictions of the architecture in Edo, and comparative secondary literature providing evidence of the city’s street morphology. The historical focus is within the third phase of the urban revolution (1715–1867); it specifically reflects the urban morphology from 1772 to 1848, from the Meiwa fire to the Kaei era. The sample area reflects the typical urban space in Nihonbashi, with archival sources dating back to the Bunka Era (1804–1808)—an era of stability.

The characteristic urban section at the Nihonbashi is one of the few examples in which we can visualize multiple levels of spatial information based on layers of historical evidence—from urban scale to architectural details—critically assess art historical research, and “depict” the lost urban space as it is lived and perceived.

With this description as a goal, this study employs space syntax methodology, employing Isovist_App (McElhinney, 2024), a free multi-platform software tool for spatial analysis. It is based on the geometric spatial unit called the “isovist,” which Benedikt (1979) defines as “the finite volume of space that is visible at any given point at which a perceiver might exist.” An isovist is typically depicted as a polygon outlined on a floor plan in two dimensions (Ostwald and Dawes, 2018) and used for analyzing architectural and urban spaces.

The essence of this approach is to uncover the behavioral response to the pattern of the built form and spatial layout as dictated by the structure of the urban space, offer a fine-grained spatial analysis and examine the embodied experience within the long-lost city streets of Edo. As a significant effort to quantify the relationship between Edo’s typical built environment and human agency, this study investigates spatial use among the streets as experienced through everyday rhythms of human agency, such as movement, co-presence and encounters (Figure 4).

The use of space syntax as a method has been widely acknowledged and has showcased its great capacity since the theories were proposed by Hillier and Hanson (1989) for cases at different urban scales. Space syntax offers an empirical reading of “the city as one thing” by mathematically modeling the relationships between spatial configuration and human agency, such as permeability, accessibility, movement and visibility (Vaughan, 2018). To date, prior studies have dealt with cities at the street scale, and scholars have developed methods for analyzing the spatial layouts of existing buildings (Pachilova and Sailer, 2020), historic neighborhoods (Lyu et al., 2023), urban areas (Griffiths and Vaughan, 2020) and entire cities (van Nes, 2021). Earlier research has also systematically described and analyzed streets, enabling measurement of the connectivity of each street space to its surroundings, how people move through cities, urban segregation and historical development, among other valid applications.

Across the globe and time at various scales, some scholars applying space syntax have highlighted the importance of investigating the building–street interface in fine-grained detail to reveal the character of private–public boundaries (Palaiologou and Vaughan, 2014). There is a growing research trend in employing space syntax to analyze past built environments (Paliou et al., 2014) and archeological spatial reality (Battistin, 2021) and sensory analysis in (urban) heritage settings (Baumanova, 2020; Lorenzon et al., 2023). Additional studies have underlined the importance of empirical observations of human movement in city streets as fundamental to understanding how city streets function (Monokrousou and Giannopoulou, 2016; D'Acci, 2019) and how the street visual experience can be improved. van Nes and Yamu (2021) reflected in their book that, through isovists, it becomes feasible to visualize sequences of visual fields along specific movement routes through urban areas and neighborhoods by providing “the geometrical area of a person’s panoptical view from a particular location in the built environment.”

There is a notable lack of application of space syntax methodologies to lost or disappeared urban sites. Nonetheless, a fundamental theme in urbanism that illuminates its complexity as a cultural phenomenon is the concept of the “palimpsest”. Traditionally, a palimpsest refers to a type of manuscript, often made of papyrus or parchment, that has been reused and overwritten, with traces of the previous text still visible. However, the term also applies to objects or locations that exhibit “diverse layers or aspects apparent beneath the surface.” (Merriam-Webster, 2024) This is particularly true for many modern urban cities, including Edo/Tokyo. These layers, whether temporal, spatial or even imaginary, demonstrate the ongoing transformation of an urban city as a perpetual palimpsest (Buckler and Hazzard, 2016). Second, while the importance of the building–street interface is recognized in the urban history research of Edo/Tokyo, there is a lack of precise quantification methods and visual analysis to examine these interfaces in fine-grained detail. Current studies often focus on qualitative descriptions, missing the intricate variations that can significantly impact urban dynamics. Thirdly, the existing body of research focuses on pure physical properties of the (lost) built environment. There is a gap in exploring other sensory dimensions, which could offer a more holistic understanding of how people interact with urban spaces.

The study’s ambition requires precise quantification of the building–street interface to examine the lost urban area in fine-grained detail, from street network configuration to building morphology. To unveil the character of the building–street interface and bring it more into reality, this research first steps into and draws a visualization of the formal boundaries of the lost built environment of Edo based on historical evidence in three major categories:

1. The cadastral boundaries show the building footprints (lots) drawn from the archival documents (Edo Kakaeyashiki Ezu, 1807; Edo Kakaeyashiki Meisaisho, 1808) from the Mitsui Bunko Archives. These documents not only depict the spatial configuration of the land owned by Mitsui but also assist in unveiling the dimensions of the alleyways and street configuration (Figure 5).

2. The urban boundaries related to the urban fabric, based on the regulations defining street widths and the dimensions of urban blocks (land), were obtained from city magistrate regulations, sectional maps from Edo Kiriezu (Takashiba and Murakami, 1854), and secondary literature investigating them.

3. Morphological boundaries are based on a bottom-up correlation of architectural styles, such as opening types and façade divisions per building typology. These are obtained from visual depictions in the Kidai Shōran (1805) (Excellent View of this Prosperous Age); works by Kitagawa (1837) demonstrating the image of the urban front in detailed urban sketches, illustrations and annotations, particularly those concerning architecture; and the highly acclaimed illustrations in Ukiyodoko (Shikitei et al., 1811) depicting the rear tenements (Figure 6).

Morphologically, this study identified three distinct types of streets (main streets, cross streets and alleyways), three types of buildings (townhouses, storehouses and rear houses) and service and communal facilities within the sample area (Figure 7). In particular:

1. The main street of Nihonbashi had a width of approximately 18 meters, as set by urban planning (Sorensen, 2002).

2. The cross street of Honshirogane-chō was laid perpendicular to the main street and had a width of about 9 meters.

3. Alleyways connected the main or cross streets to back passageways, providing intramobility within an urban block. Alleyways were officially designed to be only wide enough for one person to walk through, and thus did not result in any width greater than 2 meters, which creates a dramatic contrast in width compared to the main and cross streets. As Bestor (1985) states, they were “spatially discreet, yet at times almost invisible to the casual observer,” and even sometimes unofficially formed (Gaubatz, 2016).

4. Townhouses were prominent structures built along the main and cross streets, which created an abutted image by standing next to each other along the streets. Townhouses within the sample area possessed different frontages, from 3.64 to 5.46 meters. At street level, these houses represented the most dominant spatial components. The ground floors were mostly housing shops or workshops.

5. Storehouses were fireproof building construction systems used to store valuable goods. One of Mitsui’s lands served as a juxtaposed bookshop. Unlike townhouses, their front façades did not have a flexible character.

6. The rear tenements were low-rent housing situated in the internal parts of the urban blocks, with no direct access to the main or cross streets. There were 27 rear houses in the sample area with three different rectangular spatial configurations. They measured at the sides 3.64 by 4.55, 2.72 by 4.55, and 3.64 by 3.64 meters, resulting in a surface area of approximately 12.5 to 16.5 square meters each.

7. Service facilities for the community, such as wells, latrines, and garbage collection points.

8. Communal facilities such as guardhouses and neighborhood gates at the intersection of the two main arteries regulate people’s passage, especially at night.

The intricacy of every constructed setting relies on the physical boundaries established by communities, such as walls and doors within a structure, which impact the visibility and mobility of individuals within that space. In addition, drawing on an urban micro-spatial relationship beyond private/public and inside/outside, a hybrid level of detail is adapted at the building–street interface, with physical features represented in a two-dimensional planimetric setting. The cadastral document provides accurate spatial information on rear tenements, storehouses, townhouse boundaries and service facilities. Nevertheless, there was no evidence of openings such as doors and windows. At that point, the interpretation and building of hypotheses based on the previously mentioned visual representations, urban sketches and illustrations were necessary to meticulously map the vanished urban space in micro-detail. When dealing with lost sites, where no physical traces remain, this approach requires careful professional interpretation, logical inference, exploration of common usage patterns and visual analysis of architectural types to adapt these studies to the current site context. While this methodology is speculative compared to direct physical traces, it allows researchers to reconstruct and hypothesize about past urban forms with a reasonable degree of reliability. For instance, even if the windows of rear tenant houses might not perfectly match the hypothesized patterns, the overall reconstruction still provides valuable insights into historical urban development. Embracing this approach, despite its challenges, underscores the innovative efforts to piece together the intricate history of urban landscapes, making meaningful contributions to our understanding of lost urban areas. Although the main and cross-street dimensions aligned with the map of city magistrate regulations, there was no quantitative information on the alleyways. Extraction of the total length of the cadastral plots located on the premise of the full length of Mitsui-owned land allowed precise representation of the alleyways. This study stands out as the first to explicitly measure and delineate the dimensions of urban life at this scale, contributing valuable insights into urban morphology and spatial organization.

In exploring lost urban spaces, this study addresses how inhabitants experience visibility and couples it with how inhabitants/visitors (i.e., agents) move to quantify and spatially visualize the relationship between environmental geometry and human movement/perception by combining space syntax and isovist measures. It employs a selected set of “local” isovist measures of Isovist_App that relate to occupant experience within space (Closed Perimeter, Occlusivity, and Average Radial), “global” Space Syntax-type measure that characterize configurational relations across a plan as a whole (Choice), “semi-local” measure that spans between local and global information (Directed Visibility) and isovist agent analysis. In particular, this study explored the following:

1. How often a location can be seen from the said area by Directed Visibility (Figure 8, left panel), creating isovists to describe the space from the perspective of an agent within an environment at randomly chosen positions within the selected area. It works if someone is standing at different points within a city square and noting how often s/he can see prominent landmarks from each spot. By mapping these viewpoints and creating isovists (visibility polygons), we can understand which areas are more prominently visible and which may be hidden or less noticed from various perspectives. This shows where the features for maximum visibility and impact are.

2. The mean view ray length of all spaces visible from a location by the Average Radial (Figure 8, middle panel). Average Radial would measure the average distance of one’s line of sight to other buildings or landmarks visible from that vantage point. This metric helps in evaluating the extent of visual connectivity and openness from different locations within urban areas.

3. The unobstructed perimeter at each specific location by calculating the length of the solid and/or closed edges surrounding all visible spaces from a particular vantage point using a Closed Perimeter (Figure 8, right panel). It demonstrates as if someone is standing at a street corner in a bustling urban district and assessing and how much of her/his view is obstructed by buildings, walls, or other structures. Closed Perimeter would quantify the length of solid edges that enclose her/his field of vision, indicating how much of the surroundings feel visually enclosed or open from that specific location. This measure is crucial for understanding spatial perception and the sense of enclosure in urban settings.

4. The lengths of all open edges, which are parts not obstructed by obstacles, create a dramatic visual change as a user passes between spaces by Occlusivity (Figure 9, left panel). Occlusivity would measure the lengths of open edges between these different spaces—such as the open vista of the main street versus the enclosed alleyway. This metric helps in understanding how visual transitions occur between different urban spaces and how they influence the experience of moving through the city.

5. How often a location falls within isovists generated by random walks in space through choice (Figure 9, right panel). Choice would analyze how frequently different points in the urban space are included in isovist views generated by random walks or paths taken by people. This measure helps in identifying which areas within the urban space are more likely to be seen and accessed or restrict accessibility.

The abovementioned measures and definitions are rooted in McElhinney, 2024, which was constructed using insights from Benedikt’s (1979) groundbreaking paper and drawing upon other Space Syntax literature influenced by Hillier and Hanson (1989). As Turner et al. (2001) described, the concept’s attractiveness lies in the fact that isovists offer an intuitive perspective of spatial environments. This study describes space from an individual’s viewpoint, reflecting on how individuals perceive, interact with, and navigate through it.

Furthermore, the research also experimented with isovist agents by studying the visibility fields based on arbitrary locations and generating mass-agent trails to explore the trails (Figures 10 and 11). For visibility measures, the author used an angle of 120°, counting on the individuals’ adjustment of their gaze to view everything in the field of view.

Isovist_App applies to all spaces defined by surfaces, and could compute geometrical or relational properties as measures and display spatial representations of how such measures are distributed in space as fields. However, there have been a couple of shortcomings due to the modern architecture focus of the software and the contrasting traditional building morphology of the period: In Edo architecture, there are semi-transparent elements such as the shops’ sliding doors or the half-length shop curtains (noren). The semi-transparent elements reflect the time spans when the stores are closed and have no corresponding representation in the current software parameters. These elements cannot be entirely attributed as solid, transparent, or reflective elements in the software instructions. This study excludes half-length shop curtains and models the sliding doors as open–transparent, reflecting daytime use.

Isovist analysis quantifies the visibility of lost urban features in Edo and visualizes the area and perimeters. This study provides an intuitive understanding of this concept achieved through an innovative analytical method. The resulting measures were represented using several colored maps. In principle, they employ a color scale from red to blue, where red stands for the most visible or accessible spaces and blue represents the least visible. In other words, red denotes a high-to-view potential and blue indicates a low-to-view potential.

Assessing the visibility of locations within a given area through Directed Visibility, as depicted in Figure 8 (left panel), expounds the prominence of the intersection between the main street of Nihonbashi and the intersecting street of Honshirogane-chō. The deep blue-painted regions delineated in Figure 8 (middle panel), representing the Average Radial, symbolize the proximity of individuals to visual obstacles, such as walls. Figure 8 (right panel) employs the same scheme for Closed Proximity, wherein closed edges are defined in blue, and areas of heightened visibility potential are highlighted in red.

Transitioning to Figure 9 (left panel) of Occlusivity, dark blue hues denote spaces characterized by significant occlusion, correlating with potential limitations in mobility and exploration of proximate spaces. Conversely, the red hues in Figure 9 (right panel) accentuate zones characterized by high movement flows, indicative of heightened accessibility and traversal demonstrating Choice.

Striking contrasts among the perceptions of city streets concerning the human body are evident. Figures 8 and 9 show the divergent visibility measures between alleyways and streets. Further, Figure 8 shows the varying degrees of “movement interface” as defined by Hillier et al. (1986), and simulates the visibility of spaces for both the inhabitants of the urban block and strangers passing by. Although street widths are set normatively through urban planning, the figure illustrates how visual experience varies as one moves along the street and street façades change.

The analysis demonstrated the extensive visibility of shops at street level, where a relatively large amount of commercial activity took place, which was also visible from afar. In contrast, the residential sphere of the urban block in close proximity was highly secluded. This divergence had created a pluralistic nature in the vicinity and a prominent distinction in the embodied experience within street spaces.

Figures 8–10 show that most townhouses were highly visible in a section along Nihonbashi and Honshirogane-chō streets. The distinction between interiority and exteriority is blurred, making townhouses the dominant elements of street space. This is due to the flexible boundaries of the streets. The liminal culture of the built space was created by the level difference on which one could sit, and the degree of flexibility of the building façade created a low level of perceptual boundaries. Except for the storehouses and gatehouses in the section on Nihonbashi Street, the ground-floor spaces had similar isovist lengths. Furthermore, a comparative analysis of visibility (Figure 11) showed a lack of street surveillance at the connection of alleyways to the main street (ranging from 7 to 12 meters).

In summary, this research reveals the following insights regarding the contrasting views and encounters on Edo streets concerning the human body:

1. Visual experience and human agency: The dynamic degrees of visibility and movement interfaces along the street are striking. Figures 8 and 9 illustrate how the visual experience changes as one moves along the street and the façades of the buildings change. The analysis highlights the extensive visibility of shops on the (main) street, whereas residential areas are more secluded. The high visibility of the townhouses, as demonstrated in Figures 9 and 10, indicates the blurring of the line between the interior and exterior, contributing to a low level of perceptual boundary and a distinct embodied experience within the street space.

2. Street surveillance and community control: Investigating the concept of “eyes upon the city streets” and voluntary controls in Edo required a multifaceted approach. While gate regulations and mobility restrictions during the nights are well-documented, and the visual depictions such as the Kidai Shōran (1805) serve as rich and significant resources, the findings show how they impede life beyond the “high street.” It is fascinating to observe the stark contrast in visibility between the alleyways, where ordinary individuals such as day laborers and domestic workers reside, and the main street, which accommodates a diverse mix of people from different social classes and backgrounds. This striking contrast aligns with Jacobs' theory of street surveillance in densely populated urban areas. Notably, the visibility from the alleyways toward the main street was significantly greater than that from the main street to the alleyway. The narrow spatial culture of the alleyways, with a mean width of 1.43 meters, contributed to their seclusion, and the adaptation of the built structures to a slightly angular layout contributed to their hidden presence. However, the first 10 meters lacked any scrutiny from within.

3. Spatial togetherness and communal interaction: Edo’s densely populated backstreets likely fostered a sense of spatial togetherness among commoners, leading to a large (visual) involvement in each other’s lives. The models show that Edo residents experienced spatial proximity, which affected the social dynamics. Spatial togetherness along the back streets likely fostered a sense of familiarity and increased social interaction among commoners, contributing to a vibrant and interconnected community. In contrast, spatial configurations may pose challenges in maintaining privacy. The norms of spatial morphology dictated dense space use, mainly to promote as much economic profit as possible through rents, and shaped modularly in accordance with modular materials such as tatami mats. Despite the proximity of the rear tenements and their small size, the visibility analysis demonstrated that these houses nevertheless included a visual privacy culture; the façade organization of each house prevented seeing through the inside. In addition, the integrated evaluation of visibility and accessibility demonstrated that service facilities were carefully placed, making them spatially accessible but visually discrete.

The unique urban culture of Edo flourished and matured over more than 200 years during the peaceful government of the Tokugawa Shogunate; however, the present metropolis of Tokyo occludes the former spatiality of Edo’s street space. This interdisciplinary study has provided a vibrant empirical depiction of the lost city streets using archival resources and digital methods, and providing a fine-grain, quantitative, and visually intuitive perceptual taste of “the city as one thing.”

This study explored the translation of normative forms into built forms by examining pedestrian flows and levels of visibility, the impact of the morphological language of buildings on human interaction and accessibility measures. This has been possible by capturing multiple layers of spatial properties, bridging the embodied experience in diverse settings by examining and experimenting with mappings of further evidence, including the building–street interface, which has been fundamental to conducting a visual fine-grained urban analysis.

The sample area visually demonstrated significant urban diversity regarding seclusion when comparing houses in the front to those at the back of the urban block in Edo, representing a typical setting of the period. This study examined streets and resolved how people navigated and visually experienced them in everyday rhythms, doing so provided evidence of the close-to-real bodily experiences of ordinary people such as day laborers, domestic workers, porters, beggars and peddlers within the lost streets of Edo.

This study raises many points that require further investigation. First, extending the simulation to include a more comprehensive context within the city could foster a deeper understanding of lost streets in Edo. The results of this study are promising and should be validated through further testing of diverse case samples and cross-comparisons between different neighborhoods to extract configurational patterns in history, such as in the Meiji Era and modern times. Second, a comparison of urban spaces examining different timeframes, such as day and night or festival days, has the potential to further unveil the flexible boundaries of the space and how they affect its use. Third, the current discourse on space syntax tends to recommend elucidating three-dimensionality. Therefore, future studies should target how city streets offer an embodied experience and examine how people visually perceive spaces, such as the perception from and to the first floor. Finally, enhancing interdisciplinarity by integrating insights from sociology and environmental studies can foster deeper analyses and explore the topic more comprehensively.

This research was partially conducted as part of the postdoctoral project titled project “Digital Urban History”, within the framework of the project The Freedom of the Streets: Gender and Urban Space in Eurasia 1600-1850, under Grant Number 276-68-007, PI: Danielle van den Heuvel, funded by the Netherlands Organization for Scientific Research (NWO).

The author would like to express her thanks to Marie Yasunaga for support im navigation through archival documents and the romanization of the lot dimensions; Reiji Iwabuchi, Bébio Vieira Amaro, and Hiroshi Emoto for their help in material inquiry from the Mitsui Bunko Archives; and Sam McElhinney for his technical guidance at the experimental stages of this research. The author is also indebted to the anonymous reviewers for their insightful reviews and suggestions.