The spatial feature and use pattern of external space in Chongqing traditional urban settlement

ABSTRACT The importance of external space in residential areas is well-established in academic circles, and the spatial feature and use pattern have emerged as key aspects for relevant discussion. Research regarding the external space in urban settlements primarily involves relational holistic studies by western scholars and micro-observations by oriental scholars; both approaches have their own limitations. The research presented herein (i) studies external space as a physical space, categorizes the external space characteristics of Chongqing’s Shibanpo traditional settlements at three microscopic levels (i.e., upper elements, side elements, and bottom elements), (ii) conducts on-site observations of the use pattern of external space, and (iii) evaluates the interactions between spatial features and use patterns, as well as their joint effects on the development of the external space in the traditional urban settlement. The results indicate that various spatial elements can influence utilization behavior, and this relationship is discussed in detail to provide a foundation to guide future research and decision-making efforts.


Introduction
Urbanization is characterized by floating population and land transformation (Kuang et al. 2020). Along with the development of new housing, there are still many traditional urban settlements in China that have not been overthrown or rebuilt, and they are facing an uncertain future (Peng 2014). Currently, in the face of rapid changes in the surrounding urban environment, traditional urban settlements are finding unique ways to adapt to these dramatic changes. By relying on the original traditional living space, they are adjusting to urbanization by implementing bottom-up renewal development; this transformation is partial, minor, and continuous (Zhang et al. 2021). During this process, the target populations, lifestyles, spatial forms in the settlements all change. These changes in traditional settlements are not isolated, but rather, they comprise a complex comprehensive system in which various elements mutually influence and restrict others (Xu and Li 2021).
The external space plays a special role in settlements. The indoor space is the core space of the residence, whereas the external space provides connections throughout the entire settlement. In general, indoor spaces in traditional residential areas have difficulties meeting the needs of modern life; therefore, as a supplement to the internal space, the external space not only provides traffic connections, but also represents a place to host private and public activities (Lynch 1960). Owing to the space requirements of such functions, and to allow long-term use of the external space in traditional residential areas, people apply various simple transformations and layouts to optimize the spatial form of the external space (e.g., by adding kitchen, restaurant, or entertainment functions). Although the spontaneously renewed external spaces in traditional residential areas are typically not very beautiful or carefully designed, their use is more diverse relative to the residential areas developed from top-to-bottom ).
The external space of a residential area impacts the urban landscape and the image of residential areas, and it represents a space that residents of the traditional residential area face every day (Li 2019). The space features and the daily activities are the truest embodiment of residents' daily lives in a residential area. Therefore, understanding the features of the residential area and the daily lives of the people living there facilitates the development of traditional residential areas.
Urban external space has always been a primary focus for scholars. In the context of modernism, significant attention has been paid to the figure characteristics of external space. Trancik (1991) proposed the figure-ground theory based on the Gestalt theory, with a focus on the relative relationship between solid mass and open voids. Based on the figure-ground theory, Rowe and Koetter (1983) expanded upon the figure relationships involving texture, and considered the structural relationship of the figure-ground form as the foundation of space energy and the provision of identifiable structures. Ashihara (1979) introduced the figure-ground theory into the street and summarized the morphologies of street spaces and interfaces. Jacobs (1993) converted the western street tradition into a design paradigm in the context of modernism through case analysis based on figure relationships and environmental cognition theory.
In contrast, some scholars, such as Jacobs Jane, have objected to a pure figure spatial analysis.  challenged the God-overlook analysis governing traditional urban planning theory, and instead, shifted the perspective to that of everyday people, which elucidated the complexity of connected spatial elements and urban vitality. Alexander (1968) proposed the concept of organized complexity and regarded it as a key aspect of the construction of urban vitality. They highlighted the symbiotic relationship between urban services and facilities, as well as the possibility of spatial elements being involved in various types of activities. On this basis, Hillier (2007), Hillier and Hanson (1989) and Hanson (1999), who were proponents of Space Syntax, introduced topological thinking into the study of urban spatial networks and proposed that the relationship between space and social activities lies in the relationship between configurations of people and configurations of space. Salingaros (1998) also shifted the focus toward the relationship between space and human activities, and demonstrated that more densely connected spatial nodes and more basic elements lead to richer urban lifestyles. Research regarding urban spaces has thus shifted from figure-based to system-based. This has led to the realization that spatial vitality has emerged as an important evaluation criterion for successful cities where people are the principal elements of space, rather than a pure art form. It is therefore necessary to understand the design methods of successful cities according to peoples' daily lives.
Spatial features and activities (the core of the present research) have become two necessary aspects in studies of urban vitality, and they serve as the inspiration for this research. However, the aforementioned research represented by the Space Syntax theory abstracts the spatial units and aims to reconnect them. The approach to discussing the overall structure generally ignores differences in the basic elements, and this imposes significant limitations for regional research because the regional manifestation is more related to the microscopic material elements. The method of recording the external space at the micro-level was also emphasized in Modernology and Street Observation Theories proposed by Oriental scholars Kon(1987) and Genpei, Terunobu, and Minami (1986). However, similar to the urban planning theory that focuses on figures, the core of this approach is the meticulous recording of objects, which lacks deeper analysis and discussions of elemental relationships. Therefore, in response to (i) the limitations of a single form in studies of figure features and (ii) the macroscopic problems of systematic space research, this paper aims to explore the connection between spatial features and utilization patterns in a given spatial system at the microscopic level, thereby providing a more comprehensive perspective for regional research regarding urban spaces.
In recent years, various scholars have made similar attempts. For example, Li et al. (2000) evaluated the use of space in terms of the elements of activity occupancy, and they determined the influence of space on residents' awareness of the collective living realm. Fang et al. (2019Fang et al. ( , 2020 conducted element-level analysis of the street space in Tianzifang, Shanghai. Unlike the traditional Space Syntax theory, which divides the space according to spatial forms, their research applied the Space Syntax theory to the microscopic study of external space and creatively explored the process of establishing urban elements, element space, and a spatial network system. Zhang (2021) discussed the characteristics of street space according to basic elements and activity elements, and found that at the microscopic level, there were correlations among space elements; moreover, there was also a correlation between space elements and tourists. Inspired by these results, the present research aims to examine the systematic relationship between spatial features and activities at the micro-element level. This work therefore aims to compensate for the lack of attention paid to the micro-level by theories, such as Space Syntax theory, and to respond to the lack of attention paid to crowd activities in purely figure studies.
The external space of Chongqing Shibanpo residential area was selected as the research object. A field investigation revealed the characteristics of space constituent elements and space use patterns, and the characteristics of the external spaces in the survey area were evaluated as upper elements, side elements, and bottom elements. These elements were analyzed to explore the correlation between the external spatial features and the use patterns. The results suggested ways in which various space patterns induce certain utilization behaviors, and how the space features adapt to the utilization needs in the bottom-up development of residential areas. This research proposes a new micro perspective for space system studies, the findings provide a significant foundation for future space creation efforts and may guide space design projects that will induce positive behaviors.

Process
The external space of Chongqing Shibanpo residential area served as the research material. Field research was conducted with a focus on the spatial feature and utilization of the external space in this area. The field space was surveyed and mapped to obtain field data. In terms of space features, the overall external space was divided into upper elements, side elements, and bottom elements, which were each investigated individually. The space features of each type and the distribution of space features in the target area were then summarized. In terms of use patterns, the investigation of overflowing living items in the external space and the utilization behavior of on-site users led to the distinction of five common use patterns in the target area, namely stockpile utilization, housework utilization, traffic utilization, landscape utilization, and entertainment utilization. According to the space features of upper elements, side elements, and bottom elements, the external space of the survey area was divided into 143 units, and the space types and use patterns in each unit were counted. Finally, SPSS software was employed to conduct correspondence analysis regarding the two categorical variables of space types and use patterns to explore the relationship between the space features and the use patterns of the external space of Chongqing Shibanpo residential area ( Figure 1).

Research area and sample determination
Chongqing is located in the southwestern region of China, at the upper reaches of the Yangtze River. The terrain is dominated by hills and mountains, which account for 76% of the total area, and thus, it is known as a mountainous city (Chongqing Municipal People's Government, 2021). The survey site (Shibanpo) is located in the southern area of the peninsula where the Jialing River and Yangtze River meet, is a residential area built in a naturally mountainous environment, with typical Chongqing mountain characteristics ( Figure 2). Therefore, the external space of Shibanpo presents numerous spatial features, particularly the changes in altitude, which primarily manifest in changes in the elevation of the external space itself (the combination of steps, squares, and paths) and the diverse altitudes among buildings and the external space (e.g., stilted buildings). Based on Chongqing's mountainous environment, the upper elements of the external space also have diverse characteristics, e.g., the canopy space, the road space passing through the buildings, and the corridor space in the sky, which also influence the limitation of space. Therefore, considering these rich space features and the changes in altitude and upper shading elements, the outer space of the survey area was divided according to the convex space division theory of space syntax (Hillier and Hanson 1989); overall, 124 samples were obtained ( Figure 3) as the material for this research.

Spatial features
Considering the abundance of external space in the Chongqing Shibanpo settlement (reflected by the plane and the diversity of the entire threedimensional space), it is necessary to divide the space into multiple related components to conduct analysis at the micro-level (Jacobs 1993) and to better understand the seemingly disorderly and irregular characteristics of this settlement. Ashihara (1981) proposed that floors, walls, and ceilings are the three elements that limit a given space. He also introduced the concept of reverse space, wherein the external space is designed as an internal space without a roof, and the ground and walls are the two elements that define the external space. Owing to the large number of upper shelters (e.g., eaves, rooms, canopies) in the survey site, the upper elements were included in the grounds and walls proposed by Ashihara, and the three elements of space became upper elements, side elements, and bottom elements, corresponding to the "ceiling, wall, and floor", respectively, of the indoor space. The space characteristics of the three parts are analyzed herein to evaluate the characteristics of the overall external space ( Figure 4).
Specifically, for upper elements, there are four main spatial types: no-shading, eave shading, room shading, and canopy shading. No-shading refers to the open air, without any shading elements; eave shading refers to  the auxiliary eaves of the buildings on both sides, which extend to the upper part of the external space, thus shading the external space; room shading refers to the external space passing through the building, where the lower part is open and becomes the external space, but the upper layer is still an architectural entity; canopy shading refers to a temporary canopy device that is commonly found in the target area and is generally built with plastic film, bamboo poles, ropes, and other simple materials.
In terms of side elements, depending on the presence, absence, types, and forms of the entities beside the external space, the following six types are considered: empty, house, retaining wall, retaining wall & house (same), retaining wall & house (backward), and retaining wall & house (protruding). The empty type refers to a side that does not physically limit the external space, which is often a cliff or open space; the house type refers to a house with internal space, which often has doors and windows that are connected to the external space; the retaining wall type is also very common in the mountainous environment of Chongqing, and it refers to the artificially constructed solid wall meant to maintain the mountain wall; the retaining wall & house (same), retaining wall & house (backward), and retaining wall & house (protruding) types are three-sided types that combine elements of the retaining wall and house types. These side element types are also very common in Chongqing, thus reflecting the diverse combinations of houses and mountains in the area, and highlighting the construction activities tailored to local conditions. According to the relative positions of houses and retaining walls, these combination types are considered "same", "house backward", or "house protruding". The typical Chongqing buildings, i.e., stilted buildings, belong to the retaining wall & house (protruding) type. In this case, the external space is often limited by the side elements on both sides. Therefore, when examining the side elements of each sample area, the side elements on both sides must be comprehensively considered. There are 12 types of combinations: houseempty, retaining wall-empty, retaining wall-house, house-house, empty-empty, house-retaining wall & house (same), retaining wall & house (same)-empty, house-retaining wall & house (backward), houseretaining wall & house (protruding), retaining wallretaining wall, retaining wall-retaining wall & house (same), retaining wall & house (backward)-empty.
For bottom elements, in addition to the two common types of external spatial features (path and square), there are also a large number of steps in the survey site. The steps not only have the linear characteristics of the plane, but also have altitude differences, which are a typical external space element of Chongqing. The types of upper elements, side elements, and bottom elements of the 124 investigated samples were counted for analysis.

Use pattern
The main population in the Shibanpo area are indigenous people and tenants, since the building space is too small to meet the needs of contemporary life. Therefore, the residents use the external space according to its characteristics and their own needs (this includes privatized and public use) ( Figure 5). Based on the records of the overflowing living items in the external space of the surveyed area, and with reference to people's activities in the external space during the survey, it was concluded that private utilization in this area mainly included housework utilization, stockpile utilization, and landscape utilization patterns, whereas public utilization included traffic utilization and entertainment utilization patterns ( Figure 6). The traffic utilization pattern refers to a use pattern without overflow of living items or obvious privatization or public activities; it is thus dominated by the pure traffic function. The entertainment utilization pattern refers to the use of frequently-occurring communication and leisure activities, such as playing cards, chess, games, and other activities; in this case, the common overflowing items are leisure tables and chairs. The housework utilization pattern refers to the use of private housework, such as washing, cooking, drying, and other chores, with common overflowing items including sinks, gas stoves, drying hangers, and dining tables, among others. The stockpile utilization pattern refers to stacking sundries or idle items, which are not valuable items. Because of the limited indoor space, these items are often placed in the external space near the house. Common overflowing items include sundries, cabinets, building materials, and the like. The landscape utilization pattern refers to the simple landscape plants arranged around the households in the external space, including in the door and window areas of the house or in a private small garden. The common overflowing items in this case include various patterns of potted plants. The use patterns of the 124 investigated samples were counted, and it was determined that some of the samples had combinations of multiple use patterns. To smoothly perform the follow-up correspondence analysis in this study, the samples were separately counted; for example, Sample No. 33 involves the combination of entertainment utilization and housework utilization. Herein, it was therefore regarded as two samples, namely 33-1 entertainment utilization, and 33-2 housework utilization. Ultimately, 143 samples corresponding to a unique space type and use pattern were obtained.

Correlation analysis
Correspondence analysis is a multivariate statistical analytical method developed on the basis of R-type and Q-type factor analyses; therefore, it is also known as R-Q-type factor analysis. Correspondence analysis can reveal the difference across various categories of the same variable and the relationships between categories of different variables. It is a visualized analytical method, which can display several groups of data that seem to have no connection through a visually intuitive positioning map (Clausen 1998). Based on the spatial features and use patterns of the 143 samples collected from the spatial division and statistics described above, a cross-tabulation of the spatial types and use patterns at the three micro-levels (i.e., upper, side, and bottom) was constructed. Then, the SPSS software was employed to perform a correspondence analysis of the three levels of spatial features and use patterns. The purpose of this analysis was to obtain a corresponding map that clearly and intuitively shows the degree of correlation between different categories in a certain spatial element and the five utilization patterns.

Distribution of use patterns
The use patterns of all samples were counted to obtain Figure 8. It was determined that the types of external space utilization of the target area were dominated by traffic utilization (59) and housework utilization (43), accounting for 41.26% and 30.07%, respectively, followed by stockpile utilization (18), landscape utilization (13), and entertainment utilization (10), which accounted for 12.59%, 9.09%, and 6.99%, respectively.

Relationship between space features and use patterns
The relationships between the space features (of upper elements, side elements, and bottom elements) and the use patterns were tabulated according to the statistical data (Figure 9). Then, the correspondence analysis tool in the SPSS software was employed to perform a correlation analysis for the three sets of categorical variables to obtain three sets of correspondence maps.
In the correspondence map, a closer distance between the points of two different types of variables indicates a stronger relationship between the two. Starting from the origin point (0,0), a vector was drawn towards the two variable points of the same type. When the included angle between them is relatively small, this indicates that there is a high degree of similarity between the two variables in this set ).
In the correspondence map relating the upper element space types and the use patterns (Figure 10), the two variables of stockpile utilization and room shading are very close, indicating that they have a strong relationship. The traffic utilization pattern and the landscape utilization pattern are both close to the noshading type, indicating that the no-shading type has a strong relationship with both traffic utilization and landscape utilization. The entertainment utilization pattern and the housework utilization pattern are close to the canopy shading type and the eave shading type, indicating strong correlations among these variables. Among these, the relationship between the canopy shading type and entertainment utilization is stronger, and the relationship between the eave shading type and housework utilization is stronger. Figure 11 shows the correlation between the side elements space types and utilization patterns. It revealed that the landscape utilization pattern was close to the house-retaining wall & house (protruding), indicating that they have a strong relationship. The entertainment utilization pattern and the stockpile utilization pattern are close to the empty-empty, house-house, and houseempty, indicating that they have a strong relationship. The housework utilization pattern and the retaining wallhouse are close, as are the traffic utilization pattern and the retaining wall & house (same)-empty, houseretaining wall & house (backward), retaining wallretaining wall, retaining wall-retaining wall & house (same), and retaining wall & house (backward)-empty, indicating that they each have a strong relationship. After a vector was created from the origin point (0,0) to the variable in question, it was determined that certain side element space types (e.g., retaining wallretaining wall, retaining wall & house (backward)empty, retaining wall-retaining wall & house (same), retaining wall & house (same)-empty) have a very small angle, indicating that their similarities are very high. Figure 12 shows the relationship between the bottom elements space types and use patterns, three clear correlations were elucidated. The square is close to landscape utilization, stockpile utilization, and entertainment utilization, indicating strong relationships. The steps and the traffic utilization pattern also have a strong relationship, as do the path and the housework utilization pattern.

Discussion
Although the research object was the external space, it was discovered that nearly half of the samples have upper elements shading, i.e., eave shading (18.18%), room shading (11.89%), and canopy shading (11.19%). This indicates that the phenomenon of semiindoorization or even indoorization of the external space in the target area is common, which is in agreement with the characteristics of Chongqing's  mountainous environment. Numerous canopy shading and eave shading types also indirectly reflect the fact that the indoor space of the house is insufficient and can no longer meet the needs of users. Therefore, the users often use simple materials and techniques to expand the indoor space and transform the external space near the house into semi-indoor space.
The side element spaces comprised numerous types. The various combinations of mountains and buildings reflect the awareness of users to build their homes based on the local conditions. Based on the degree of openness, this can be categorized as twosided empty, one-sided entity with one-sided empty, and entities on both sides. The one-sided entity with one-sided empty (i.e., house-empty, retaining wallempty, retaining wall & house (same)-empty, retaining wall & house (backward)-empty) account for approximately half (53.14%) of all samples, followed by those with entities on both sides (39.17%) and those that are empty on both sides (7.69%). The spatial features of side elements exhibit diversified spatial distribution with combinations of openness and closedness, which are also typical of Chongqing's mountainous residences. Based on the types of entities, they can be divided into combinations of house, retaining wall, retaining wall & house (same), retaining wall & house (backward), or retaining wall & house (protruding), and 12 types can be obtained overall. The side element spaces that were originally extremely diverse and even a little messy were thus clarified through classification. However, it is worth noting that since the scope of this survey focused on the external space that was accessible to people, Chongqing's landmark architectural type (i.e., stilted buildings) or other retaining wall & house combinations, do not constitute the main part. This is because the spatial features of the stilted buildings mainly appear on the front of a hillside or cliff, and the scope of investigation in this chapter is mainly concentrated in the interior of the residential area.
The number of steps in the bottom element space is related to the other two conventional forms (paths and squares), which also represents a key difference between the external space of mountainous settlements and plain settlements. The most basic function of bottom elements is traffic. The various combinations of steps, paths, and squares create rich walking experiences in the external space of the investigated region.
In terms of the use patterns, in addition to the basic traffic utilization, more than half of the space is used for housework (30.07%), stockpile (2.59%), landscape (9.09%), and entertainment (6.99%). This indicates that the use of external space in the investigated region is relatively dynamic and diversified. The housework utilization pattern and the stockpile utilization pattern represent life logistics, among which, the housework utilization corresponds to a relatively high proportion.
Specifically, the users often use the external space as a drying area, a laundry space, a kitchen, or a dining room. There are also some landscape-type and entertainment-type utilization activities. This indicate that although the residents belong to mainly low-income groups, in a living environment with very limited resources, they still create simple leisure places for spiritual entertainment.
Finally, the correlation between the space features and the use pattern is discussed. In the correspondence map of the upper element spaces and the use patterns, three clear clusters are observed, which reflect the degree of indoorization. The upper room shading is mainly a result of the bottom of the building entity being opened and converted into an external space. They have the highest degree of indoorization, privacy, and security, which lead to a high correlation with stockpile utilization. The upper canopy and eave shading types are closer to semi-indoorization. The semi-indoor space creates favorable conditions for housework and entertainment activities. For example, the semi-indoor space can protect people from the rain, but it is well ventilated and thus suitable for use as a kitchen. In particular, canopies are often made of translucent materials, so they have a higher degree of openness than the eaves. This may contribute to the stronger relationship between canopy shading and entertainment utilization. The upper no-shading space has the highest degree of openness and the strongest relationship with both traffic utilization and landscape utilization. This may be because traffic utilization is entirely attributed to public use, whereas the upper shading actually forms somewhat of a private space. Therefore, the traffic utilization is more closely associated with unoccupied open spaces without shading.
It is clear that in addition to the retaining wall, the side elements of retaining wall & house (same) and retaining wall & house (backward) are associated with the traffic utilization pattern. The key characteristic distinguishing the retaining wall from the house is that there is no connection between the internal building space and the external space, and no one enters the external space from the inside. Therefore, the external space corresponding to the retaining wall has low vitality and low richness of utilization; therefore, traffic utilization is the main pattern. The retaining wall & house (same) and retaining wall & house (backward) comprise a house; however, because of the difference in height relative to the external space, they often have no direct internal-to-external connection. As a result, these are very similar to the retaining wall type. The house-retaining wall & house (protruding) has the attributes of stilted buildings and a strong relationship with landscape utilization. This may be because the stilted buildings often have protruding balconies or corridors. Potted plants are commonly placed at the interface between these spaces and the adjacent external spaces for daily appreciation. As a result, entertainment utilization and the side element space have a strong relationship, indicating that people like to conduct public entertainment activities in a highly open space. Interestingly, the housework utilization pattern and the house-retaining wall type have the strongest relationship, indicating that families living in this type of space enjoy a high level of privatization in the external space; in the external space with houses on both sides, users will make concessions because of competition in the field and will not excessively privatize the external space.
In the correspondence map of the bottom element spaces and use patterns, it is clear that the square has a strong relationship with landscape utilization, stockpile utilization, and entertainment utilization. This may be because the squares generally have larger areas, giving them high potential for developing into spaces with composite functions. The step type has a strong relationship with traffic utilization, which is consistent with the diverse topographical characteristics of Chongqing's mountains. The step is one of the major traffic elements connecting spaces with height differences. In the conventional understanding, the path type should be used primarily for traffic utilization; however, the statistical results indicate that the path type also has a strong relationship with housework utilization. This indicates that there is no appreciable change in the height difference, thus allowing a large number of users to conduct private housework activities in the access space. This is also indicative of the privatization of external space in the investigated region.
In addition, the impact of the microclimate environment on the creation and utilization of space was investigated. Because of the hot summer weather with high humidity in Chongqing, the external space adopts a comfortable microclimate environment and becomes an important place for community and family life. Such spontaneously-formed spatial forms adapt to the local climate environment (especially the upper elements), satisfies the ventilation needs, and avoids the adverse effects of both the scorching sun and rainfall. These upper elements and the abundant side and bottom elements jointly create a rich external environment and promote specific activities.

Conclusion
Based on site investigations, the external space of the Chongqing Shibanpo residential area was surveyed and mapped to identify the overflow of living items and record the use behavior of residents. The external space features of the target area were first sorted according to the three aspects of upper elements, side elements, and bottom elements. Based on the data regarding space features and use patterns, the correlations between the space features and use patterns were evaluated. The state of the external space created spontaneously in the residential area was thus revealed, and the ways in which the users adapted to the natural terrain to create usable living spaces in the available external spaces were discussed. The specific conclusions are as follows: The external space of the target area comprises rich forms and characteristics, including upper elements, side elements, and bottom elements. In particular, the shading of rooms, the retaining walls of side elements, the combinations of retaining walls and houses, and the steps of bottom elements are all unique features of the exterior space in Chongqing's mountainous residential areas. Semi-indoorization is common, reflecting the limited amount of internal space and the trend of expanding one's living space.
Users have a high degree of enthusiasm for the use of external space; both private and public use types are diverse. With the exception of the basic traffic utilization pattern, a large amount of external space is used for life and logistics activities, and a small amount of space is used for entertainment.
There is a strong relationship between the morphological characteristics of the external space and its utilization. Upper shading determines the degree of indoorization of the space, which impacts the utilization tendencies. For example, a space with a high degree of indoorization is often used for closed stockpile utilization, and a space with a moderate degree of indoorization is used for housework and entertainment, whereas an entirely outdoor space is generally used for traffic and landscape appreciation. The side elements determine the plane limitation of the external space, the openness in the horizontal direction, and the connection between the side space and the external space (i.e., whether there is a connection between the building interior and exterior, and whether the accessibility of the external space is high or low). These aspects influence the ultimate utilization of external space. The bottom element space determines the movement and staying tendencies of users in the external space. The access-type space with significant linear characteristics has a strong relationship with housework utilization in the mountainous settlements of Chongqing, whereas the steps serve a pure traffic function. These relationships are distinct from those of the spaces in plain residential areas. The upper elements (degree of indoorization), side elements (openness in the horizontal direction), and bottom elements (movement and stay) jointly contribute to the domain attributes of the external space, thereby affecting users' behavior when occupying the space. In summary, a higher degree of indoorization formed by the upper elements, a stronger indoor-outdoor connectivity formed by the side elements, and a lower passability formed by the bottom elements, will lead to higher domain attributes of the space, which ultimately facilitate the private activities of surrounding users. However, spaces with opposing attributes will likely induce public use, such as traffic. The special relationships between these space attributes and utilization behaviors result in a rich community life in the external spaces in the Chongqing urban settlement.
The strategy of microscopic observation that is proposed in this research is innovative and fills a key knowledge gap in past research. Furthermore, the results demonstrate that the space of bottom-up development is regular, rather than chaotic. The discussed space system provides a combing method and a strategy for elucidating complex spatial systems. This research also investigates the details of the important relationship between spatial features and space utilization, which can be used to guide future external space design.

Prospects
This research explores the correlation between spatial and utilization types at the micro-element level and draws valuable conclusions that can serve as references for future designs of external space. However, the tendency and guidance of the type are fundamental, and therefore, to further define the effects of spatial features on utilization, it is necessary to quantify the spatial indicators (e.g., area, length, width, etc.) and utilization indicators (e.g., duration, intensity, etc.) in subsequent research.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Notes on contributors
Lin Chen is a Research Associate, Department of Architecture, Waseda University, Tokyo, Japan. He received master degree of architecture in 2017 from Waseda University. He is currently an Assistant Professor at architecture department of Waseda University, FURUYA Nobuaki Laboratory, Japan. He is interested in the historic districts preservation and regeneration.
Kai Fang received doctoral degree of architecture in 2020 from Waseda University, and now is an Assistant Professor in Nobuaki Furuya Laboratory of Research Institute for Science and Engineering, Waseda University. He is interested in the modern creation and extension of traditional Chinese design.