A micro-scale study on the spontaneous spatial improvement of in-between spaces in Chinese traditional districts considering the relationship between modifications and encroachment

ABSTRACT As the transition between streets and houses, in-between spaces represent unique and important residential components in East Asia, which reflect the spontaneous spatial improvement awareness of users in traditional districts and the state of public-private relations. To fill a gap in prior research regarding in-between spaces after spontaneous spatial improvement in traditional districts, this study considered 97 houses on Jiazhi Street and evaluated the two expressions of spontaneous spatial improvement of in-between spaces by quantifying modification indicators and encroachment using field surveys and area statistical methods. The research examined the correlation between various modifications and encroachment using multiple regression analysis and evaluated the characteristics of spontaneous spatial improvement by users based on multi-group relationships. The results offer a new perspective for observing the changes in residential environments in the context of urbanization, thus providing a basis for developing renewal strategies for traditional districts in China. GRAPHICAL ABSTRACT


Introduction
Indoor and outdoor spaces have been separated since humans began to build shelters (Nooraddin 1998), and this situation is exacerbated in modern residential areas. The relationship between buildings and outdoor spaces is often overlooked or the two areas are considered completely separately (Schittich 2012) (Hanson 2000), meaning that open spaces are often ignored (Can and Heath 2016). Numerous scholars have recognized the importance of indoor-to-outdoor connections, and proposed concepts, such as in-between spaces (Nooraddin 1998), interfaces (Hillier and Hanson 1989), translucent spaces (Furuya 1996) (Furuya 1999), soft edges (Gehl 1986), and elastic spaces (Wang 2020) to complete this spatial relationship (Can and Heath 2016). In this research, the transitional space between a building and the street is defined as the in-between spaces, analogous with the concept proposed by various scholars, including Nooraddin and Van Eyck (Nooraddin 1998) (Van Eyck 2008. An in-between space refers to the space dimension between indoor and outdoor; it is not a facade, but rather, it constitutes a volume (i.e., with thickness) between a building and the street. This dimension is somewhat ambiguous, or even variable. In the present study, the protrusions of the facade and the encroachment in the front space define this dimension ( Figure 3). As a transitional space between inside and outside, the space between a building and the street comprises a boundary between publicness and privateness, and its ambiguity leads to diverse forms and functions. Although such areas are relatively flexible, uncertain, and sometimes even considered to be strange (Grosz 2001), they are also associated with identity characteristics and have rules, which are not conducive to randomness (Wolfrum 2018). In-between spaces can inspire positive social interactions and behavior, as well as friendly neighborhood relationships (Can and Heath 2016). Furthermore, they represent an important component of urban spaces because they can create an environment for social-spatial relationships, pay homage to historical causes (Wang and Cao 2020), and reflect the local living environments, policies, and reallife conditions (Nooraddin 1998). However, a wide range of problems facing contemporary urban development are reflected in such boundary areas (Yuan 2010). Therefore, research regarding in-between spaces can provide a unique perspective for understanding the relationship between cities and houses, while offering ideas for improving urban spaces.
In-between spaces reflect the local construction style and culture, and have their own unique styles in East Asia (represented by China), especially in traditional districts after spontaneous spatial improvement. Since the 1980s, China has experienced the fastest urbanization process in the history of the world (Guan et al. 2018) (Chen, Liu, and Tao 2013). Nobel Prize winner Joseph Stiglitz regarded China's urbanization as one of the two major events affecting human progress in the 21st century (Stiglitz and Yusuf 2001). In addition to the top-down unified development, the spontaneous spatial improvement of traditional districts is often the first step to improving the local living environment owing to the associated low costs and technical requirements ; as a result, this strategy is becoming an alternative to building new houses in metropolises (Plaut and Plaut 2010). Spontaneous spatial improvement refers to the autonomous decision-making regarding housing modification activities to improve local living environments, wherein individuals represent the decision-making units. This is a varying bottom-up model that lacks unified planning (Lu and Jiang 2012) (Tan, Tang, and Guo 2019) (Zhang 2016). However, despite the lack of professional designers, spontaneous spatial improvement is by no means accidental (Rapoport 1988). In fact, this strategy has a close relationship with residents, reflects real user needs and living conditions, and follows a set of rules (Lu 2009). Since Bernard Rudofsky first paid attention to the aesthetics of the shelters built by people spontaneously (Rudofsky 1987), people began to realize the continuous impact of spontaneous spatial improvement activities on the living environment, which can represent real urban spaces (Kudva 2009).
Owing to their ambiguous flexibility, in-between spaces provide an opportunity for spontaneous spatial improvement; both are uncertain and flexible, and both follow simultaneous logic and regularity. Therefore, the spontaneous spatial improvement of inbetween spaces is particularly prominent in China. As Jane Jacobs described (Jacobs 1993), the problem of the city is an orderly and complex problem; the spontaneous spatial improvement in the in-between spaces seems chaotic, but in fact, it is closely associated with local life, culture, and technology.
To date, there have been significant research efforts involving in-between spaces. Can and Heath (2016) investigated the impact of in-betweens spaces on social interactions via space syntax analysis and snapshot observations. Zerouati and Bellal (2020) examined the relationship between the permeability of outdoor in-between spaces and social interactions. Hickman (2013) identified the importance of external spaces in poor areas to encourage residents to participate in community interactions. Van Nes and López (2007) found that the micro-scale spatial relationship had a significant impact on urban street life. Gehl, Kaefer, and Reigstad (2006) pointed out that the specific characteristics of outdoor intermediate spaces impact the communication between residents. Wang (2020) discussed the influence of boundaries between residential areas on urban spatial vitality. Hatsuda (2007, 2009) investigated the store-front planning in Japan before and after the Great Kanto Earthquake, and found that the architectural and commercial perspectives had distinct concepts of storefront spaces.
Numerous researchers have evaluated spontaneous spatial improvements and proposed relevant concepts, including self-organization (Cilliers 2002) (Portugali 2012), informal urbanism (Bromley 1978) (Mukhija and Loukaitou-Sideris 2015), and self-help housing (Jimenez 1983) (Sanya et al.2012). Plaut and Plaut (2010) analyzed the impact of personal, household, location, and neighborhood variables on such modifications. Lu (2008) (Lu 2009) proposed that spontaneous construction was the driving force for the formation of regional buildings, and (Wolfrum 2018) explained the law of spontaneous renovation of residential facades. Junzo, Momoyo, and Yoshiharu (2001) collected information on various unnamed and demand-oriented buildings in Tokyo, which were formed spontaneously and provided the best economic benefits and social value. Zhang et al. (2021) discussed traditional street spaces following spontaneous transformation and their close association with spatial elements. Fang et al. (2020) analyzed the new structural characteristics of the street system after informal space reorganization from a microscopic point of view. Chen et al. (2021) discovered that the degree of street activity and residential remodeling rate exhibited a negative correlation. They also confirmed the importance of empowering students to understand informal spaces (Mukhija and Loukaitou-Sideris 2015).
Previous research has primarily focused on either the buildings themselves or only the externa l space, while generally ignoring the connection between inside and outside. Although there have been numerous macro-scale studies regarding in-between spaces and spontaneous spatial improvement from the perspectives of urban morphology, social economy, and culture. Scholars in the field have reached a general consensus regarding the role of in-between spaces in terms of promoting public life. Designers and managers understand that in-between spaces should be improved to stimulate and support a vital and vibrant community. However, Chen et al. (2021) concluded that the modification of buildings does not necessarily promote the vitality of the street, and Yashiro and Nasu (2005) found that the alley space and the building interior become closer and more dynamic with longterm modifications. This demonstrates that scholars have inconsistent views on the relationship between spatial modifications and activity encroachment. The present study suggests that the reason for this inconsistency is that they are not simply describing a relationship between two concepts, but rather, a relationship between sets of multiple feature groups. In previous studies, there is limited analysis of users' spontaneous spatial improvements of in-between spaces based on observations and establishing a complete multi-factor association system at the micro scale. As a result, such studies cannot provide direct or explicit guidance regarding realistic spatial design strategies. Academic research into the phenomenon of spontaneous spatial improvement of inbetween spaces and a deeper understanding of users' spatial behavior and logic will help clarify the complex relationship between users and these types of spaces. The insights gained can also remind designers of the attitudes and demands of users towards this type of space, so they can improve the design of such spaces in the future.
This research focuses on the spontaneous spatial improvement of in-between spaces in traditional districts. The traditional district of Jiazhi in Zhejiang Province, China, served as the research object, and the modification and encroachment of in-between spaces are discussed herein. The research investigated how the users improved their living environment through daily spontaneous spatial activities in Chinese traditional districts in the context of urbanization by analyzing 97 samples. The results provide a unique perspective regarding spontaneous spatial improvement. The conclusions from this study improve our understanding of the spatial characteristics and daily use in this type of district. Therefore, this paper represents an important reference for the decision-making required for traditional district revival strategies.

Materials and research procedure
Jiazhi is an ancient fishing town located on the south bank of the Jiao River in Taizhou City, Zhejiang Province, China (Fang 2019). According to ancient books, including the "Jiading Chicheng Records", Jiazhi has been an important terminal for the collection and distribution of local waterborne goods since ancient times. In the middle of the 18th century, when commercial ports gradually began to flourish, it once again emerged as a collection and distribution center (China Taizhou Net 2013). Following urbanization, newly-built residential areas have become people's preferred type of residences. Representing one of the most prosperous areas in Taizhou, Jiazhi has lost its former glory, and it has instead become a cheap area for migrant workers and local people who cannot afford new houses Figure 1 (Fu 2018).
The buildings on Jiazhi Street, which served as the survey subject, were constructed in the middle of the Qing Dynasty. They are primarily residential buildings facing the street, with beautiful details and the forms of traditional wooden buildings in southern China. In response to the long-term residential demand and user changes since urbanization, a large number of spontaneous spatial improvements have emerged. Meanwhile, the public and private boundaries between houses and streets have become blurred. This is a typical case of spontaneous transformation in a traditional district, which is commonly found in East Asian cities (Miao 2001) (). Upon investigating the field houses one-by-one and excluding the newly constructed buildings after complete demolition, a total of 97 houses along the street were selected as the sample for this study. Their in-between spaces have undergone rich modifications, and numerous daily activities take place here (Figure 2).
In this study, modification refers to artificial changes to the physical elements of the building for maintenance or functional purposes (e.g., replacing doors and windows, installing canopies and other spatial transformations), which will have relatively long-term spatial impacts. Encroachment refers to placing temporary (movable) items outdoors for the user's daily activities, thus encroaching on the in-between space, e.g., placing tables and chairs outdoors for meals, playing mahjong, placing sinks outside to wash. The encroachment may change daily and is more flexible. The former changes the spatial form, while the latter, as an expression of domain consciousness, is affected by space. Both spatial activities reflect the users' awareness of the living environment, and the relationship between these aspects should be discussed. In the inbetween spaces of the target area, almost all spontaneous spatial activities are characterized as either modifications or encroachments. The research presented herein discusses modifications and encroachments themselves, while regarding them as two spatial expressions of users' awareness of space quality improvement. A holistic analysis was conducted, and their characteristics are summarized, and their relationship is discussed herein.
The in-between spaces of 97 houses were surveyed, the planes and facades were mapped, and the building components and overflowing encroachment items were recorded (Appendix 1). An area statistics method was employed to quantify the modification and encroachment aspects. Finally, seven modification indicators were identified as independent variables, and the encroachment was used as the dependent variable. Multiple regression analysis was performed to evaluate the correlation between various types of modifications and encroachment activities to further reveal the characteristics of spontaneous spatial improvement of inbetween spaces in traditional districts (Figure 3).

Modification analysis
Modifications occurred primarily in the physical components. The modifications of in-between spaces in the survey area were concentrated in facade modifications and entrance ground modifications. As the boundary between the inside and the outside (Askari 2009), the facade comprises the dual spatial attributes of buildings and streets. Although it is an architectural element, the facade affects the external space. Facade modifications were categorized as wall modifications, additional modifications, and doorwindow modifications for this analysis. During the site survey, the facade of the building was mapped, and the modified sections were recorded. Considering that most of the buildings in the surveyed area were two-story buildings, they were subdivided into ground-floor facades and upper facades to distinguish the differences in the impact of different levels of facades on the in-between spaces. The modification rate of each type was equal to the modification area divided by the facade area of the floor. Finally, a wide range of facade modification indicators were obtained through statistical analysis, including the wall modification rate of the ground-floor facade, the wall modification rate of the upper facade, the additional modification rate of the ground-floor facade, the additional modification rate of the upper facade, the doorwindow modification rate of the ground-floor facade, and the door-window modification rate of the upper facade ( Figure 4).
The ground element of in-between spaces represents the connecting area between the inside and outside, and it is also one of the public-private boundary elements (as is the facade) (Saito, Tatsumi, and Almazán 2019). The ground element provides visual connectivity and has the function of allowing for body passing. Therefore, the form of ground elements and user improvements are additional aspects affecting research on in-between spaces. There were numerous ground modifications in the surveyed area. Herein, we summarize three basic elements, i.e., platforms, steps, and ramps, to describe the ground modifications of 97 samples. Similarly, an area statistical method was employed to quantify the ground modifications. According to the quantification method of Zhang et al. (2020), the ground modification rate is equal to the ground modification area divided by the bay width of the building (Figure 4).

Encroachment analysis
The public-private relationship between streets and buildings has become blurred in East Asia (Ashihara 1979) (Miao 2001), mainly because residents place objects outside in their daily lives, and many daily activities occur in external spaces; as a result, private objects and activities overflow into public spaces. The boundary between the public space and the private space is therefore ill-defined and flexible (Wang 2020), thus representing a very typical spatial phenomenon in the target area. Because of the difference in overflow objects, the in-between spaces that originally had no definite function are given a flexible function that originally belonged to the interior of the building, i.e., the kitchen, laundry room, living room, dining room, etc. These features comprise an informal space; even without architectural elements, the furniture, objects, and resulting human activities define the space.
The quantification of encroachment objects can represent the users' awareness of spontaneous use of space (Wang et al. 2019). Common quantification methods include statistical quantity (Fujitani and Kobayashi 2017) and statistical area (Wang et al. 2019)  analyses. Considering the significant differences in the size of encroachment objects, the statistical projection area approach was employed in this study to attain greater accuracy. According to the plan obtained from surveying and mapping, the projected area of the encroachment objects of the 97 samples was calculated. Similar to  the ground modification rate, the encroachment rate is equal to the projected area of the encroachment object divided by the bay width of the building ( Figure 5).

Correlation analysis
Finally, the wall modification rate of the ground-floor facade, the wall modification rate of the upper facade, the additional modification rate of the ground-floor facade, the additional modification rate of the upper facade, the door-window modification rate of the ground-floor facade, the door-window modification rate of the upper facade, and the ground modification rate were used as independent variables, and the encroachment rate was used as the dependent variable. Multiple regression analysis was conducted with Statistical Product and Service Solutions (SPSS) to evaluate the correlation between various modification indicators and encroachment.

Modification
Among the 97 samples, only three had no facade modifications, and only three had no ground modifications. The average facade modification rate was 0.36, and the average ground modification rate was 0.55. These values indicated that modifications are very common in the target area, and the degree of modification is relatively high. Among the various types of facade modifications, the door-window modification rate was the highest (0.24), followed by the wall modification rate (0.08), and the additional modification rate (0.04). This may be because the traditional wooden doors and windows cannot meet the safety and natural lighting requirements of modern life, so they have a high probability of being replaced. There were also variations in the facade modifications on different floors. In general, the facade modification rate of the ground-floor was significantly higher than that of the upper floor. This indicated that the demand for facade improvement was primarily concentrated on the ground-floor, which also confirmed that the facade of the ground-floor plays a more significant role as a medium for indooroutdoor interactions. However, the additional modification rate of the ground-floor facade (0.02) was lower than that of the upper floor (0.05). This is because air conditioners and billboards are mostly hung on the upper floors (e.g., air conditioners are usually set up in the bedrooms on the second floor). Additional elements with identification attributes (e.g., billboards) are often positioned on the upper facade to enhance their visibility (Appendix 2). The platform element is the most common ground modification (88), followed by steps (50), and ramps (21). The platform is for static interactions, while the steps and ramps are for passing by. Ramps have been added in recent years because many users have electric motorcycles, which require ramps to facilitate access (Appendix 3).

Encroachment
Only nine samples showed no encroachment, so the occurrence rate of encroachment reached 90.7%, and the average encroachment area reached 2.4㎡with a 0.62 encroachment rate. This indicated that the privatization of in-between spaces by users was very common, and a significant number of activities occurred in those in-between spaces (Appendix 2).

Correlation between modifications and encroachment
Using Casewise Diagnostics and applying three times the standard deviation of the standardized residual as the criterion, it was determined that the values of E-18, W-18, and W-19 were abnormal, and their standard residual was greater than three (Table 1). With the site inspection during the investigation period (which was during the Chinese Spring Festival), these three houses had an abnormal quantity of overflow goods related to the centralized temporary sales of Spring Festival commodities by vendors; this was quite different from the typical living situation (Figure 6). Because these abnormal values caused deviations in the results, the analysis was performed again after removing these outliers. Table 2

Discussion
The spontaneous spatial improvement of the traditional district is a phenomenon that cannot be ignored in the process of urban development. The in-between spaces, which comprise the transition between the street and the house, are the main areas where spontaneous spatial improvement is expressed by users when facing issues in the public and private fields. Scholars have investigated the importance of in-between spaces to the residents' public life, but most of the reports take macroscopic perspectives, and therefore, it is difficult to provide clear suggestions for spatial practices at the specific operational level. Based on these viewpoints, the present study established a systematic model to describe modifications and encroachment during spontaneous spatial improvement at the micro scale. Furthermore, we have clarified the mechanism of influence among micro elements in the in-between spaces via multiple regression analysis. The correlations between modifications and encroachment, as two expressions of users' spontaneous spatial improvement, were thoroughly demonstrated.
In contrast to the conclusion drawn by Chen et al. (2021) that the activeness of space usage is negatively correlated with residential modification rates, the present research involved a more detailed investigation and demonstrated that the modification type and modification location have distinct and complex relationships The walls of the ground-floor facades in the study area were mostly wooden structures, which changed to solid brick walls (which were stronger and safer). These walls separated the interior from the exterior, generating a negative attitude towards the utilization of in-between spaces. As a result, spontaneous encroachment is uncommon. Additional modifications of the upper facade were the only changes related to the encroachment among all upper facade modifications. This is likely because the interactive markers (e.g., billboards) are usually positioned on the upper floor, reflecting the user's understanding of internal and external interactions, which are also the driving force for encroachment. The additions on the upper floor serve to attract people to conduct activities in the in-between spaces. This result, along with Tang's and Ding (2015) concern about the relationship between facade signs and street dynamism, calls into question Ashihara's (1981) argument against secondary contours formed by architectural additions.
Interestingly, door-window modifications on different floors had the opposite effect, and this distinction was related to the differences in the properties of indoor and outdoor connections of doors and windows between different floors. Gehl (1987) pointed out that when people move outdoors, the field of view is shifted downward by approximately 10°, thus drawing focus to the ground floors. The interior of the ground-floor can be connected to the activities on the street through the doors and windows on the first floor. Similarly, the encroachment in the in-between spaces can be seen clearly from indoors (e.g., whether a bicycle parked at the door has been stolen, whether the clothes being dried have been blown away by the wind, etc.), and people can also easily reach the outdoors (e.g., when people see neighbors passing by the door, they can go to chat). Moreover, users who are active in the in-between spaces can easily know what is happening indoors; for example, users may enjoy the fresh air in the in-between spaces while observing the children who are studying indoors at any time. Therefore, the door-window modifications of the ground-floor facade are related to the users' awareness of the joint use of in-between and indoor spaces. Doors and windows that connect the interior with the exterior promote the outdoor encroachment. The upper floor's doors and windows are separate from the street owing to their height. Thus, door-window modifications of the upper facade do not promote the use of in-between spaces, but rather improve the indoor living environment of the upper floor. This indicates that users pay more attention to the indoor environment, while ignoring the outside. Zhang et al. (2019) and Arima et al. (2008) similarly suggested that the transparency and openness formed by door and window openings connect the inside to the outside, thereby enhancing their interactions. However, the present study considered the relationship between these openings and encroachment on different floors in more detail. It was determined that the upper doorwindows indicate the opposite result, which is a breakthrough considering previous research. In addition, Tang and Ding (2015) found that the higher the position of the sign on the facade, the more attractive  it is to the flow of people. The possible reason for this observation is that the higher the sign, the better it can be seen. However, the purpose of doors and windows is to connect the indoors and outdoors, not to let people see the doors and windows themselves. Therefore, the present research showed the doors and windows in the ground-floor can promote people's encroachment activities in the in-between spaces. Although Zhang et al. (2021) proved that some architectural elements impact the encroachment elements, they ignored the key role of the ground in the in-between spaces in terms of encroachment. The present research found that ground modifications (e.g., adding platforms, steps, and ramps) privatized the street space. These ground elements represent a user's declaration of ownership of the in-between spaces. The separation of in-between spaces from the completely public streets makes these areas ambiguous, which promotes encroachment by temporary objects. If the original encroachment overflows to the street considering the wall of the facade as a starting point, it can now further overflow to the street based on the ground element. This strengthens the excuse for users to expand the scope of encroachment.
In contrast to previous studies focusing on the relationship between space modifications and encroachment, as well as research indicating that in-between spaces promote social interaction, the present work not only confirmed the related macro relationships, but also went further to explain the relationships among various elements of modifications and encroachment in the in-between spaces at a more specific (micro) level. This study indicated that different elements presented different results, which can provide direct suggestions for future space design.

Conclusions
Previously, it has been common to ignore the connection between residences and the city when studying the residences alone or the urban space individually (Yuan 2010). Therefore, this study considered the Jiazhi Street in Taizhou, Zhejiang Province, China as a case study, and focused on in-between spaces to investigate the transitional space between the street and the houses in the traditional district following spontaneous spatial improvement. This study examined the relationships between various types of modifications and encroachment in the in-between spaces using diverse methods, including field surveys, quantitative statistics, and correlation analysis, then discussed the phenomenon of users' spontaneous living environment improvements in traditional districts in the context of Chinese urbanization.  In many developing countries, low-income families cannot afford new houses, so they have to find ways improve their existing houses (Jimenez 1980) (Strassmann 1984). The research presented herein discusses the widespread modification and encroachment occurring in the in-between spaces of traditional districts. The results demonstrated that in general, users have a strong awareness of spontaneous spatial improvement and street space privatization, and they can better meet the needs of daily life through simple modification methods and temporary prop settings. These activities strengthen the connection and interactions between the interior and the exterior.
The results indicated that there are differences in the user space awareness regarding different modifications and how they impact in-between spaces. The wall modifications and the door-window modifications of the upper facade reflect the user's understanding of attaching importance to the internal space, in contrast to the active use of in-between spaces. The additional modifications of the upper facade, the door-window modifications of the ground-floor facade, and the ground modifications all strengthen the interactions between inside and outside, and expand the scope of privatization, with the same driving force for space improvement as the territorial encroachment; they are mutually positively correlated. Previous macroscale research could not provide direct guidance for spatial practices; however, the results presented herein offer a breakthrough and will undoubtedly provide advice to designers and authorities regarding specific elements to they can focus on the elements that are positively related to the encroachment of outdoor activities and apply them to further enhance the spatial vitality of the in-between spaces. In the future, with the improvement of research models and the expansion of survey sites, this method can provide an important basis for designers and authorities to formulate design strategies.
Benedetto Croce (1923) proposed that all history is contemporary history. Traditional districts continue to play a role in the contemporary context of population movement and land expansion brought about by urbanization. This research verified that users adjust the contradictory properties of in-between spaces in terms of isolating or connecting the interior and exterior through the spontaneous spatial improvement of in-between spaces; these also help them adapt to the needs of modern life. The results further revealed that spontaneous spatial improvement should be viewed and discussed as a systemic issue comprising multiple problems, and such considerations provide a new model through which to discuss urban living environments.
There are several limitations of the work presented herein. First, the research model involving modifications and encroachment was developed based on the specific research site. To apply this method to in-between spaces in different regions, it should be adjusted and optimized according to the actual situation in the site to form an applicable research model. Second, the present study recorded the on-site modifications and encroachment through modernology and used the results to suggest the users' awareness based on the analysis of on-site data. However, empirical investigations of user opinions are lacking. Future investigations should aim to explain users' motivations for activities in these spaces and determine how pedestrians (i.e., users of public space) view the encroachment behavior of the owner of the house. Such insights will contribute to a more effective reference for formulating space development strategies that can simultaneously meet the needs of different user groups.