Mapping subjectivities through interviews : A design workshop in a Japanese provincial town

295 words) Workshops have become a common practice in the academic world of architecture and in participatory community building. During the workshop implementation, interviews to local residents is one of the fundamental phases. This paper develops a mapping technique to visualize the results of a series of individual structured interviews, taking as study case a workshop realized in a Japanese provincial town (Ichikawamisato) with the aim of revitalizing its urban core. The paper makes a detailed description of the research method and the visualization technique. It also includes a discussion about the generalization of the kind of qualitative research conducted in this study through the concepts of analytic generalization and case-to-case transferability of knowledge.


Background
The design workshop has become a common educational practice in schools of architecture and urbanism, and also a tool to achieve consensus in community planning, as in the so-called charrette workshops developed in USA, UK and Australia (Smith 2012:2).During these design workshops, participants − either architecture students led by their instructors, or professionals − develop proposals on a particular site in a short period of time of usually one week or less.
Typically, design workshops include a period for survey or fact finding, often based on fieldwork, in order to clarify the physical characteristics of the site.Increasingly too, these surveys address social or behavioral aspects related to the inhabitants or users of the site.This attention to the social dimension of space, which might seem essential to the lay person, was nevertheless largely neglected until the early 1960s, when a systematic body of research emerges as a critique of the predominant visual approach to urban design in Modernism, and its perceived failure to support the social dimension of the built environment.
This new body of research, called "the social use" tradition (Jarvis 1980), or "urban life studies" (Gehl and Svarre 2013) has its pioneers in Lynch (1960), who turns to examining the mental image of the city; Jacobs (1961), who made detailed observations of everyday use of public space to build her critique of modern planning; Alexander's Pattern Language (1977), where he describes the relation between specific spatial designs and its positive effect for the social interaction and well-being of its users ;Gehl himself (1987) in Scandinavia, and Whyte (1980) in New York.
Most of the methods employed in the urban life studies have been adopted from the behavioral sciences, which divide methods into quantitative and qualitative (Mertens 1997).Compared to quantitative research, qualitative research has three characteristics (Merriam 1998:6): firstly, quantitative research takes apart a phenomenon to examine component parts, while qualitative research can reveal how all the parts work together to form a whole; secondly, in qualitative research the researcher is the primary instrument for data collection and analysis; thirdly, qualitative research usually involves fieldwork.Due to the short and intense character of the design workshop, it is often difficult to use quantitative methods, which require a large sampling and statistical processing, and qualitative methods are often more suited.
Within qualitative research, the methods can be divided into observation, interviews and document reviews (Mertens 1997).Gehl and Svarre's research (2013), for example, is based on direct observation through several techniques of counting, mapping and tracking.These techniques require extensive observations, in different periods (weekdays and weekends), seasons, and weather conditions, to find reliable patterns.Document reviews are another information source, but typically the available documentation is prepared by the workshop organizers, and in many cases the questions to be identified during the workshop have not been explored yet and there are not enough documents published on the particular topic of the workshop.
The interview method offers comparatively at least two advantages.On the one hand, it is a feasible way to achieve plenty of information in a short time.On the other, it offers the opportunity for the workshop participants to get in contact with the locals and understand directly the situation for which they will have to offer design proposals.Interviewing has indeed become a common tool in workshop methodologies (Murata 2001, Kobayashi 2002, Brandt 2006).Kobayashi (2013:15-16) describes charrette workshops as consisting of six phases: 1) Sharing basic information (understanding the goal and process) 2) Site survey (understanding the area through a site visit) 3) Collecting local opinions (interviewing residents) 4) Analysis of the site (extracting "problems" and "potentials") 5) Specific design for an effective site 6) Presentation to the public, recording and dissemination.
In this paper we will focus on a method to interview residents to collect local opinions, and how to visualize them in order to share the results among the workshop participants and the public.

Research goal
Our first goal is to develop, put into practice, and test a method to realize structured individual interviews and visualize the results graphically, as part of a design workshop realized in the town of Ichikawamisato (Japan).Sec-Mapping subjectivities through interviews: A design workshop in a Japanese provincial town Gaku Inoue, Moe Kusano, Jorge Almazán.DOI 10.14198/i2.2015.03.06 ondly, to examine the validity of these results beyond the specific site (Ichikawamisato), and by doing so, to explore the possibility of building a design workshop theory, which can grow as a field of research and production of knowledge by the accumulation of different cases of workshops.

Previous research
This study is based on several bodies of research.Firstly, it acknowledges and aims to incorporate the "social use" tradition (Jarvis 1980), also known as "public life studies" (Gehl and Svarre 2013) into the practice and theory of the design workshop.
Also, it recognizes the theories of the charrette workshop, a practice that has also been introduced to Japan and has become a common tool in the so-called machizukuri (literally "town making") movement.Especially relevant for this paper is Kobayashi's (2013) long-term research and practice, who applies the charrette workshop practice in the provincial town of Takahashi in Japan.
In terms of methodology, the third body of research examined comprises mind mapping techniques.These can be considered to have started with the situationists' psycho-geographies (Debord 1955), and Lynch's (1960) cognitive maps.While the situationists addressed the individual impact of urban space, Lynch aimed to clarify shared patterns in the recognition and legibility of physical aspects of cities. Applied to cases in Japan, similar ways to map subjectivity have been developed by Yamamoto et al. (2005), and Oishi et al. (2006), in which participants freely chose urban spaces and their preferences and choices have been represented graphically.Sepe (2009) also performed similar subjectivity mapping fieldwork in Barcelona, and Almazán et al. (2012) for the city of Alicante.
Finally, this paper also employs the affinity diagrams method to synthesize qualitative data, also known as the KJ method, named after the initials of his inventor, Jiro Kawakita.Originally devised for cultural anthropological research, it is a bottom-up approach for understanding data (Kawakita 1996).As a research method, it has been recognized by the Architectural Institute of Japan (Ueno 2005).Affinity diagrams can be used in two ways: as a consensus-making tool, or as a method to synthesize qualitative data.In the former way, they are often used in management and planning processes to achieve inter-subjective consensus amongst the participants (Brassard 1989, Scupin 1997).Specifically for consensus making in an urban design workshop it has been employed by Almazán et al. (2013).The second way to use the method, as a way to synthesize qualitative data, is usually implemented by few researchers, as in Ushino (1995), who employs it to visualize and organize the opinions of residents in regional surveys, or Taniguchi et al. (1980) to summarize opinions of municipal staff in charge of urban planning.In this way it has been also used to organize architects' statements to understand the historical flow of architectural thinking (Okuyama et al. 1994).In these three later cases, researchers (typically two or three) create the affinity diagrams by themselves and include the records of the raw data in the research publication to ensure the confirmability of their investigation.We used the method in this second aception, as a method to synthesize the qualitative data from the interviews.

Relevance of the research
As explained before, interviewing in design workshops is a common practice.This paper provides a step-by-step description of a method to visualize interviews, and discusses its validity, providing a possible framework to build a design workshop theory.We consider that the growing polarization of the architectural profession into researchers and practitioners needs to be revised, not by denying the importance of neither of those two activities, but by creating a bridging theory through which the practice of the design workshop can be linked to theory and systematic knowledge.We expect to contribute to this task through the particular case of interviewing and visualizing for design workshops.

Basic information
The authors conducted the interviews in the town of Ichikawamisato (Yamanashi Prefecture, Japan) as part of a broader design workshop to revitalize its urban core (Fig. 1 and Fig. 2).The authors have been engaged in the town revitalization since 2013, when they started having contact with local community groups.After several meetings with these groups (Fig. 3) we identified the necessity of a better understanding of the residents' opinions about the urban core or central districts in order to identify problems and potentials.Ichikawamisato suffers a situation similar to many others provincial towns in Japan.On the one hand, there is a continuous hollowing of the urban cores: the disappearance of a lively, compact center, in favor of a dispersed, car-dependent life style.Town centers have become gradually empty, small shops have disappeared, and big-box stores proliferate in the cheap land surrounding the city (Nakade 2003:10).On the other hand, the increasing economic disparity between big and small cities forces the young population to emigrate to larger urban areas in search of opportunities for education and employment, accelerating the depopulation and aging (Nakafuji 2008:30, Sagawa 2008:10).
We acknowledge the difficulties to address complex problems like depopulation shift or unemployment from the point of view of urban and architectural design.However, based on the compact city theory (Dantzig 1973, Rogers 1997) we argue that an enhancement of physical designable aspects, such as density, use diversity, or accessibility to public transport, can be decisive to mitigate or even reverse the consequences of urban sprawl and provide an all-age lively urban environment.Having this goal in view, and with the support of local community groups, we embarked on an exploration of the citizens' opinions about the existing urban core to identify places with problems and potentials.

Methodological step 1: Interviews
The interviews were conducted as structured interviews by six groups.The structure of the interviews was prepared and tested by the authors in advance, and each group was instructed how to fill and record the interviews.
The six groups, formed by a mixture of university students and local high-school students, realized the interviews on-site on May 4 and 5, 2013 to the citizens of the urban core of Ichikawamisato (Fig. 2).Table 1 shows the details of the survey and group composition, and Table2 the details of the composition of the persons interviewed.

Roles within each team Question
Keio University students 7 ・Interviewing: person in charge of the conversation.・Recording: person in charge of video or sound recording.・Fill-in: person in charge of writing the interviews and marking the places on the map.With the collaboration of local community groups, letters were sent to the residents to announce the dates and contents of the survey, and to kindly ask for their cooperation.The multiple tasks of interviewing, video recording, mapping and filling the questionnaires were done by the university students.However, the collaboration with local high school students proofed to be essential.In front of young members of their own community, the residents were more open to respond.The interviews were conducted in private houses and shops, and in the streets.All requirements to ensure the ethical treatment of data were guaranteed.In the interviews we asked the residents to choose 3 to 5 "good places" and 3 to 5 "bad places" in the above-mentioned urban core.We asked them to indicate the places in a map and we marked the location in front of the interviewees.Additionally, we asked them the reasons why they consider those places good or bad.

Step 2: Affinity diagrams
The results of the interviews were synthesized qualitatively through affinity diagrams, a method consisting of making groups of categories for verbal data.To guarantee the confirmability of the results the literal statements from the residents or raw data have been recorded in

Place with a distinctive feature
This park is related to the Kabuki./ I went to the park to know about the Kabuki./ I often go to the park to visit exhibitions./ There is a peony garden./ There is a castle./ I go to the park to see theater.

Abandoned place
There was no castle in this town, so the castle is meaningless and making an economic loss for the town.Place with bad atmosphere The park's image is bad./ There are no specific things in this park.

Dangerous place
There are rumors that a ghost can be seen in this park.

Place name Reason category Raw data Haba Park Place with good atmosphere
There are cherry blossoms./ We can get a whole view of Kofu basin.
Attachment place When I was a child, I used to go with friends.Place with a distinctive feature

It is nice view place. / There are ancient ruins. Not lively place
There are no people.
Town Ground Place with good atmosphere There are poplar trees.

Around Town Ground
Convenient place Close to my home

Place with a distinctive feature
There is a monument of fireworks./ There is bell rings.A couple will be happy, when ringing the bell.

Lifeless place
There are no people./ There are people, Only when there is an event. Seishu

Bad atmosphere place
There is a bad atmosphere.

Dangerous place
Car speeds are high.

Lifeless place Desolated space
Place with bad atmosphere It's hard to become friendly with foreign people.The way of life style is different.We obtained several general categories to explain the preference for places considered as "good" or "bad".In order to clearly show the results of the analysis henceforth we will show the resulting categories in pointing double angle quotation marks («»), and the raw data in quotations marks (" ").For example, we included positive descriptions of places considered as good under categories such as «attachment place» (represented in statements such as "I used to go this school", "When I was child, I used to go to this park") or «lively place» ("I go to the shop to meet my friends", or "Sound of children's voice make me happy").As for bad places we used categories such as «abandoned place» ("The hospital should be rebuilt" or "There are few street lights") or «lifeless place» ("There are no people" or "There is no station staff").

Step 3: Mapping
The places considered as good and bad have been mapped with gradients of transparency to show the frequency of replies (Fig. 4).The following map (Fig. 5) represents the places selected by two or more people and the reasons obtained from the affinity diagrams as icons.This resulting map (Fig. 5) combines quantitative information (the number of people who selected one particular place), with qualitative data (the reasons for preference).

Results
The overall results have been summarized in Fig. 5.This map shows the rich information obtained through the interviews and represents a visualization of the multiple subjectivities of the residents.In this paper, we would like to make a more focused description of the results by focusing on the top five most liked and top five most disliked places (Table 4).4).The four parks are located in the outskirts of the urban core, and the Konjaku-dōri Shopping Street is located in the center of the urban core.
The Konjaku-dōri Shopping Street received the higher number of selections with categories describing it as a bad place.This is a typical case of a shopping street in the urban core of a provincial town, which has lost its liveliness.Almost all shops along the street have closed and there are few people on this street.This street has been selected as bad place through the following categories: «abandoned place» ("Few street lights", "The street is narrow and the ambulances can't enter", "When I walk with my child, I feel narrowness"), «place with bad atmosphere» ("There is no attraction").However, this street is nevertheless considered as an «attachment place» ("In the old days, there were a lot of old stories", "I used to walk around"), due to the memories of citizens as an old main street, and also as a «convenient place» ("Easy to buy because there are a lot of shops").
As for the four parks, they share similar features.All of them are located at the periphery, and have been selected as bad places because they are «abandoned places» ("Disordered", "This park is making an economic loss for the town", "There are few street lights").However, Hirin Park and Kabuki Park appear also in the top most liked places, under the positive categories of «attachment place» ("I go to the park often", "When I was child, I went to the park", "I started working in this park, so I like this park") and «place with a distinctive feature» ("There is a flower garden", "There is a castle", "There are monuments of calligraphy").This shows the existing controversy about these parks, and that a definite opinion does not emerge from the results.

Most liked Places
As for the top most liked place, we obtained six places: Kabuki Park, Mitama Hot Spring, Hojuin Temple, Selva Supermarket, Ichikawa High School, and Hirin Park.From these six places, three are located at the periphery of the urban core: Kabuki Park and Hirin Park (explained above), and also the Mitama Hot Spring.This Hot Spring is one of the most popular spots but does not provide opportunities to revitalize the urban core in a walkable, compact city manner since it is far from the urban core and only accessible by car.
Three places are located in the urban core: Hojuin Temple, Selva Supermarket, and Ichikawa High School.The Selva Supermarket is typologically a big-box store.This building typefree-standing, single-floor structures sitting in the middle of a large parking lotis often found in the periphery of Japanese provincial towns.Usually part of large chains, they offer relatively low prices that eliminate the competition of the small family-run shops in the Mapping subjectivities through interviews: A design workshop in a Japanese provincial town Gaku Inoue, Moe Kusano, Jorge Almazán.DOI 10.14198/i2.2015.03.06 town.Selva Supermarket, nevertheless, is located inside an urban block of the city core, and along the shopping street.Highly accessible on foot for the community, its location inside a block keeps the continuity of the urban façades.It is a rare case of big-box store integrated in the fabric of the urban core (Fig. 6).Selva Supermarket was selected as a good place by the following categories: «attachment place» ("I used to go everyday"); «place to play and shopping ("Good for sport", "Good for shopping"); «convenient place» ("Close to my home", "I often go to buy daily things"); and «lively place» ("I go to the shop to meet my friends").
Fig. 6 Axonometric view of Selva Supermarket (elaborated by the authors) Hojuin Temple was selected as a good place by the following categories: «place with good atmosphere» ("There is cherry blossom", "The view from temple is beautiful", "There is a peony garden"); «attachment place» ("When I was child, I often play in this temple", "There is my family's tomb", "I ring the bell to inform the time"), and «place with a distinctive feature» ("There is a garden made by a historical famous gardener", "This is historical temple", "The mother of Kukai (Buddhist priest) was dead in this temple").
Ichikawa High School was selected by the following categories: place with good atmosphere ("There are beautiful cherry blossoms"), «attachment place» ("I used to go to this school", "I want to keep this high school", "I like this high school's wall"), and «lively place» ("This high school's clubs are thriving", "Sound of children's voice make me happy", "My friend goes to this high school").

Discussion on the results
As stated above, the research goal of this paper is double: first, to put in practice and test a method to map subjectivities; secondly, to examine if the knowledge produced can contribute to theory building.We have also exposed our concerns about the polarization between "professionals" and "researchers" in the field of architecture and urban design.This polarization is caused by the increase in the standards and complexity of disciplinary expertise, but also by the increasing demand from universities and funding agencies to produce "objective" knowledge in a clear, systematic and measurable way.We aim to contribute to the development of research protocols that reconcile the seemingly opposed traditions of "objectivist" science and "subjectivist" design.
Mapping subjectivities through interviews: A design workshop in a Japanese provincial town Gaku Inoue, Moe Kusano, Jorge Almazán.DOI 10.14198/i2.2015.03.06 The specific question that arises here is: How to achieve the two goals of this study (subjectivity mapping, and validity for theory building) by examining a single study case?The "subjectivist" disciplines of design are based on the study of highly regarded case studies or "masterpieces".These cases conform a "canon" that all designers study, know and ultimately, try to overcome (Eisenman 2008).On the other hand, the "objectivists" natural and social sciences often consider case research as not appropriate to produce generalizable knowledge.Our position bridges this opposition: we argue that the design disciplines can produce new knowledge beyond the canon of masterpieces by trying to find ways for possible generalization of their practices, like the practice of the design workshop explained in this paper.
Following Firestone (1998) we will describe two approaches for the possible generalization of case studies and qualitative research.One is the analytic generalization where "the investigator is striving to generalize a particular sets of results to a broader theory."(Yin 1989:44, cited in Firestone 1993).This generalization to a theory means "to provide evidence that supports (but does not definitively prove) that theory" (Firestone 1993:17).
Firestone describes another type of generalization: the case-to-case translation, which coincides with what Lincoln and Guba (1985) describe as transferability.Mertens (1997: 355) calls attention to the quality criteria for this type of research: "With this approach, the burden of proof for "generalizability" lies in the reader, and the researcher is responsible for providing the "thick description" that allows the reader to make a judgement about the applicability of the research to another setting."In this paper, we have provided a detailed description of our method in order to guarantee this transferability, i.e. the transferring of knowledge on methods and conclusions to other cases by other researchers.
We will discuss now the results of this study from the above-mentioned two points of view: analytic generalization and transferability.

Analytic generalization
First, from the point of view of the analytic generalization, the results of this investigation can be considered in the three groups below in relation to existing urban design theory.
-Results that provide evidence to current urban theory.The importance of the traditional shōtengai (the commercial street) has been widely recognized in urban studies (Sasaoka 2002, Kokubu 2007, Arata 2012).This paper provides further evidence of its importance.The shōtengai of Ichikawamisato was one of the most selected places, both for good or bad reasons, which shows the mixed perception on the street: on the one hand, the memories of its vital past are still fresh, on the other hand, its obvious decay has created a negative image.
-Case-specific results.Some results can only be understood from a local and place-specific point of view, and do not seem to provide relevant evidence for a broader design theory building.The strong dislike of the Hirin Park, for example, is partly due to the local perception that it was a bad investment from the beginning and the park is making a continuous economic loss for the town.
-Results that suggest new evidence for theory building.We can also identify results that are neither confirming established existing theory nor exclusively site-specific in character, i.e. results that can suggest ways for theory to be expanded.The case of the Selva Supermarket is the most clear.The common conflict in provincial towns like Ichikawamisato is the gradual hollowing of the shōtengai or shopping street in favor of big-box stores at the outskirts which can only be accessed by car.Selva Supermarket offers an in-between solution: It is a big-box store, but since it is located in a void space within a central urban block, it keeps the continuity of the street space, and it is accessible on foot for a majority of residents.The interviews reveal the role of the Selva Supermarket, not only as a shopping place, but also as a meeting place.This case can be generalized as a an example of "cross-scale coupling": an imaginative way to combine two urban scales -the small, walkable urban core, and the large system of highways and big-box storesin a new integrated typology.The Selva Supermarket in Ichikawa provides evidence for this possibility of cross-scale coupling.

Transferability of the research
This paper provides a "thick description", a detailed step-by-step explanation of all phases implemented in the research.In that sense, we aim to keep the transferability of the methods and conclusions contained in this paper.However, we believe that a critical reflection by the authors, as summarized below, can also help not only to transfer results to other studies but to improve them in future research.
-On Mapping: We found the mapping results (Fig. 5) as a useful way to summarize a multiplicity of data in one single image, both for the workshop participants and for the public.Maps, however, are best appreciated in large prints, and therefore mapping techniques have limitations when published in the small-sized media of journal papers.The incorporation of digital formats, with interactive figures to zoom-in and out, could be a promising development to integrate large and complex visualizations into the academic format of the journal papers.
Mapping subjectivities through interviews: A design workshop in a Japanese provincial town Gaku Inoue, Moe Kusano, Jorge Almazán.DOI 10.14198/i2.2015.03.06 -Disappearance of critical cases.The method of affinity diagrams needs to be applied consciously, without discarding discordant opinions, small but relevant nuances, and critical cases.These deviant views can have a great potential in urban design to discover neglected places with urban potential.During this research we found ourselves often reviewing the categories in order not to lose these deviant views, which can provide innovative points of view.
-Empathic connection.The interview method showed extremely positive to develop empathic connection with local residents.Workshop participants often showed an increased motivation and engagement after having direct contact with the locals.We consider this contact and engagement a key condition to develop the necessary sensibility to the residents' lifestyles and values, in order to develop design proposals.
-Participatory method.The short and intensive character of the workshop requires team work.We perceived that conducting the interviews in groups was also very helpful for the group cohesion, and for the quality of the interview itself.

Conclusion
This paper develops a method to visualize in a single map a multiplicity of opinions of a population (Ichikawamisato) gathered through structured individual interviews.In spite of the apparent specificity of the results, we argue that from our study we can contribute to a wider knowledge from two points of view: analytic generalization and transferability.The analytic generalization confirmed some known problems found in Japanese provincial towns, and suggested the principle of cross-scale coupling that could be applied in other towns.As for the transferability of knowledge, we gave a detailed description of the study and we added critical reflections on the method for future improvement of the method.Particularly, as suggested in the discussion section, we are concerned with the danger of disappearance of critical cases in the method.How to keep the criticality of design disciplines and the potential for disputing generally received opinion through polemical transformation is one of the pending challenges that need to be addressed in further studies.
With this paper we have attempted an approach to research in architecture and urban design that reconciles two traditions: On the one hand the "objectivist" tradition of natural and social science and its paradigm of clear, systematic process of inquiry.On the other hand, the "subjectivist" tradition of fine arts, humanities and design, based on learning from masterpieces or relevant study cases.Specifically, we hope that this study can encourage the research community to document, analyse, and discussing multiple and specific workshops in order to collectively develop a "design workshop theory".

・
Age and gender ・Good places in Ichikawa or around the town, and the reason (Min 3 places, Max 5 places) ・Bad places in Ichikawa or around the town, and the reason (Min 3 places, Max 5 places)

Fig. 4
Fig. 4 Frequency of replies about good and bad places (elaborated by the authors)

Table 1
Details about the interviews (realized on May 4 and 5, 2013) (elaborated by the authors)

Table 2
Composition of the interviewees (elaborated by the authors)