The relationship between education and child welfare in Japanese children’s self-reliance support facilities

ABSTRACT In comparison to the foster care system in western countries, in Japan most child protective care is conducted in facilities. This article examines the institutional changes necessary for the introduction of school education in children’s self-reliance support facilities (CSSF) and considers the relationship between the spheres of education and welfare from the narratives of the school teachers who conduct the education practices and the facility staff who conduct the welfare practices. By examining these issues, this paper seeks to identify the challenges in introducing school education to the CSSF and discuss how the spheres of education and welfare can work together to overcome them. Following the introduction of school education into the CSSF, study guidance in the facilities was taken out of the hands of the facility staff and was conducted by qualified teachers. From teacher and welfare staff narratives, this introduction of school education was seen as an ‘erosion by school education’ for child welfare. However, from the perspective of Goffman’s total institution approach, this erosion has gradually generated new practices that differ from the conventional perspectives of school education or welfare. Building off of this perspective, this article suggests that school education in the CSSF can allow the children to temporarily experience a pseudo-society separate from their life in the child welfare facility, and this can offer benefits for the child’s eventual social re-integration.


Problem identification
To encourage self-reliance in children under protective social care, child welfare facilities and school education systems need to collaborate as school education has been found to play an important role in the socialization of children. Collaboration is necessary not only for everyday social care but also for the child's education after leaving the welfare facility. In a recent Japanese study on children who needed special support, judicial studies, pedagogy, and child welfare were jointly explored (Fujikawa, 2011). Concretely, these studies clarify how a good relationship between schools and family or welfare facilities is constructed. However, the practices that govern education and child welfare are different as they are controlled by different ministries and require different licenses. Even when both education and welfare are dealing with the same children, their understanding of the child and the methods of practice are different, leading to a conflict between the provisions required for education and those required for child welfare.
This study addresses the question, 'What are the problems arising from this conflict between the education system and the child welfare system?' and explores how practitioners can work together to overcome this conflict. To that end, this paper focuses on the children's self-reliance support facilities (CSSF) (jidō jiritsu shien shisetsu) where school education was introduced under the Child Welfare Law (Jidō fukushihō) revision of 1997. The CSSF has been selected because its unique combination of welfare services and particular educational needs makes it necessary for both welfare and education practitioners to consider their differences in logic and practice as school education is introduced into the facilities.

Child welfare in Japan
Compared to Europe and America, where foster care in foster families is more common for state wards, wards of the state in Japan are generally cared for in child welfare facilities (see Figure 1).
As Figure 1 shows, the rate of foster care in Japan is the lowest of all advanced countries. Human Rights Watch (HRW), a nongovernment human rights organization, found that Japanese child consultation centers that determine the treatment appropriate for children have tended to prioritize parental will, with most parents wanting their children to be taken care of by a welfare facility rather than by foster parents (HRW, 2014). Further, the 'Guidelines for the Alternative Care of Children,' voted for in the United Nations, states that the 'removal of a child from the care of the family should be seen as a measure of last resort and should, whenever possible, be temporary and for the shortest possible duration' (United Nations 1996, 'General Assembly', Article 14).  (Ministry of Health, Labour and Welfare, 2014).
From this perspective, therefore, Japanese care for wards of the state in child welfare facilities goes against international norms. Facility care in Japan has received significant criticism, primarily because children are being deprived of the chance to be nurtured by society and because the children have to live in closed isolated spaces away from society. Facility care, however, is still conducted in many countries; therefore, it is possible that facility care can have positive outcomes (Courtney, 2009). Further, it is necessary to assess whether Japanese facility care does in fact prevent children from nurturing their social natures by examining the actual conditions of the facilities.
In Japan, social protective care for state wards is offered in five different types of facilities: infant homes, foster homes, short-term therapeutic institutions for emotionally disturbed children, maternal and child living support facilities, and the CSSF. In most child welfare facilities, the children are separated from their families and are housed in communal living (except for the maternal and child living support facilities, in which the mother lives with her child). Each communal living space is a closed space separated from the society and in the CSSF in particular, the degree of isolation is higher than in any other child welfare facilities, although to a lesser degree than the Shōnenin (Juvenile Training School) which are under the control of the Ministry of Justice.

Outline of the CSSF
The CSSF, which is the focus of this study, is prescribed in the following way by the Child Welfare Act article 44: A children's self-reliance support facility shall be a facility intended for admitting children who have committed, or are likely to commit, delinquencies and other children in need of daily life guidance, etc. due to their family environment or other environmental reasons or having those children commute there from their guardians, and providing necessary guidance to those children in accordance with their individual circumstances and supporting their self-reliance, as well as intended for providing consultation and other assistance to those who have left there.
The reasons for admission are diverse but are generally because of juvenile delinquency (theft, runaways, or injury) or activities or events in the child's domestic environment (child abuse or sexual violence from parents). Therefore, the support facilities take care of children who do not receive help at school or from their families. Eighty percent of children in these facilities are junior high-school students sent by consultation officers or children who were put into protective custody by the family court. The personnel at these facilities offer: (1) life guidance, (2) work guidance, and (3) study guidance. These facilities seek to ensure that the children's lives are stable and that they have the ability to be independent after leaving the facility. In Japan, there are 58 such facilities that are home to 1500 children.
When school education was introduced into the CSSF due to the revision of the Child Welfare Act in 1997, significant changes were necessary. In other child welfare facilities, the right to education is guaranteed, with most children attending a school close to the facility. In contrast, in the CSSF, education is conducted in the branch school inside the facility and the children go to school without exiting the facility; that is, school education is embedded within the welfare system.
After the revision to the act, licensed teachers were invited to conduct the education classes, reducing the task load of the CSSF facility staff and giving the children access to higher quality learning. However, because school teaching practices that are different from the welfare staff in terms of the legal foundation, views on children, and license, have been embedded into the child welfare facilities, the child welfare facilities have been disrupted. This has generated conflict between the education providers and the welfare providers and has also been exacerbated by the over one-century-long history of the CSSF. In the next section, the introduction of school education into the CSSF is discussed.

Context
Following the stipulations in Clause 1, Article 26, of The Constitution of Japan ('All people shall have the right to receive an equal education correspondent to their ability'), it follows that the children in the CSSF have the right to receive education as citizens of Japan. Article 48 of the Child Welfare Act states: [A] children's self-reliance support facility and a foster parent shall send to school the children admitted to those institutions or entrusted to those foster parents as if they were the guardians of such children as provided in the School Education Act.
This means that the CSSF children are obligated to attend school, an obligation that was stipulated in the revision to the Child Welfare Act in 1997. Before this revision, children at CSSF facilities received 'study guidance' as 'equivalent education,' which was not the same as the regular education curriculum.
In the Kyōgoin (Reform School), the predecessors to the CSSF, educational practice was institutionally regulated as 'equivalent education.' The obligation to conduct public education had not been imposed on the reform school heads in the Child Welfare Act in 1947 until its revision in 1997. Therefore, up to this time, the equivalent education had been conducted by facility staff and was aimed at 'removing delinquency.' This long tradition of equivalent education had been the reason why school education was not introduced. Further, as the practices behind equivalent education and school education are vastly different, many reform school practitioners found the introduction of school education in reform schools to be extremely disruptive.
What is the difference between equivalent education and school education? I discussed this controversy with a principal from Kinugawa Gakuin, Naotaro Kojima, who agreed with the introduction of school education, and with the principal of Tankai Gakuen, Noboru Ishihara, who disagreed. Kojima and Ishihara had taken contrasting views regarding the introduction of school education into Kyōgoin at the conferences of principals held in 1963 and 1965. According to Kobayashi (2006), who analyzed the controversy at these conferences, Kojima had argued as follows: The object of juvenile reform is juvenile delinquency, and it is important to guarantee the right to receive education for children on the assumption that they are juvenile delinquents. But it is not desirable that study guidance be considered the only way to the Child Welfare Act to impose the obligation of public education on the head of reform schools is necessary. (Kobayashi, 2006, p. 56) However, Ishihara opposed the introduction of education on the following grounds: Why should the reform school follow school education, considering that the recent education system is in a serious state and causing juvenile delinquency to increase? I think that Article 48 of the Child Welfare Act is an excellent letter of the law because it grants great freedom to the educational practice in reform schools. (Kobayashi, 2006, p. 57) In summary, Kojima argued for compulsory school attendance for the children due to their guarantee to the right to education. In contrast, Ishihara argued that the CSSF need not have regular school education as schools generate juvenile delinquency. Although both of these interlocutors were CSSF principals and worked in child welfare, their opinions about the introduction of regular school education in CSSF were quite different. Therefore, the question is whether young juvenile offenders need or can cope with regular school education practices.
The ongoing controversy was brought to an end with the introduction of school education into the facility as it was decided that the reform school children were being disadvantaged because the equivalent education was not conducted by qualified teachers. However, the introduction of school education caused various problems; for example, how were they to register the children in school. While children were able to obtain an approval to 'delay or exemption from ordinary school attendance' under Article 18 of the School Education Law, they would lose their former school register, which would make it difficult to authorize their graduation from an ordinary school. Given this, most children had been authorized for graduation by returning to ordinary schools as they approached graduation. Before the revision of the Child Welfare Act in 1997, it was possible to issue a certificate of completion for the equivalent education. However, no reform schools issued these certificates as it was a disadvantage to the children's futures if they had a certificate that stated they had attended a reform school (Kobayashi, 2006, p. 12).
Another problem was that it is quite difficult to create a school report for the high school entrance examination. Although school reports are usually created by ordinary schools, assessing the equivalent education conducted in reform schools was quite difficult. Therefore, children received assessments by taking the routine tests from ordinary schools, submitting artwork done in the reform school, or playing an instrument in front of teachers at ordinary schools. However, Kobayashi, who has worked at a reform school, pointed out that there were also cases in which all the children received the worst grades as it was impossible to conduct assessments (Kobayashi, 2006, p. 13).
To eliminate these various disadvantages for children who received a 'delay or exemption from public school attendance' within the equivalent education system, school education was stipulated in the 1997 revision of the Child Welfare Act. In general, facility staff supported the principle of equivalency in education after the introduction of school education. However, it has been difficult for the teachers and facility staff to establish a mutual understanding. In the next section, I review the existing research on juvenile delinquency and the relationships between education and welfare to provide a brief overview of how these fields have been discussed so far.

Existing research
Most Japanese studies on children who need special support have primarily been in child welfare, pedagogy, and the administration of justice fields. Child welfare studies in particular have been conducted on issues related to juvenile delinquency. Takayoshi Doi found that delinquent behavior resulted from poverty and a difference in consciousness from that in more supportive middle class families (Doi, 2012). Tokoro also found that social neglect was a form of child abuse (Tokoro, 2014). A 'social exclusion' concept proposed in France has been adopted to analyze children who have dropped out from child welfare institutions (Iwata, 2013) and a private safety net called the 'Children's Shelter' was implemented to protect children unable to receive public support (Konishi, 2016).
In contrast, pedagogical studies have mainly focused on children with disabilities, with some focused on juvenile delinquency, child welfare, and the administration of justice. In this context, pedagogical research has examined juvenile delinquency from an education perspective. Hirota analyzed the transformational process of children in juvenile training schools with a focus on the educational environment (Hirota, 2013). Tsushima found that juvenile delinquents tended to maintain their relationships with other delinquent peers from the judicial facility when they entered the local community (Tsushima, 2013) and suggested that sending delinquents to judicial facilities did not reduce the relationships with local delinquent peers. Although these studies have focused on actual situations within facilities and the transformation of the children placed in them, there have been few studies that have examined cooperation between school education and welfare services in the branch schools established in these facilities.
In Europe and North America, studies on child welfare have focused on the development of foster children, with most demonstrating a common awareness of the behavioral or psychological problems experienced by foster children because of their separation from 'family care' (Cohen, 2005;Dozier, 2001;Fuller et al, 1996;Jones, 1985;Kolivoski et al., 2014;Ku and Plotnick, 2003;Masten and Wright, 1998). Nonetheless, foster care in Europe and North America has also been criticized in a similar way as Japan for a lack of family care. There have also been some studies focused on the relationship between education and welfare. Conger and Finkelstein (2003) found that more than half the children who had to change schools to be cared for by foster parents tended to have a poorer record at school, primarily because the foster parents and caseworkers did not inform teachers and other school personnel of the children's custodial status due to the concerns about them being stigmatized by the foster care label. Based on these findings, this study stresses the importance of cooperation between spheres of education and welfare.
As described above, research in Japan, Europe and America has discussed how education and welfare cooperation could lead to better outcomes. Building from these insights, but with a critical eye to the dynamics on the ground, this paper focuses on the problems occurring due to the conflict between the education system and the child welfare system and examines how both education and child welfare practitioners can overcome this conflict.

Methodology
Overview of field observation Fieldwork was conducted for four days each week at CSSF X from June to September 2014. The children were observed at the branch school at CSSF X in the morning and at the institution's farm in the afternoon. Semi-structured interviews were conducted with seven teaching and five welfare staff members (see Table 1), in which questions were asked about the issues and events observed during the fieldwork and the transformations in education and welfare perspectives that may have occurred among the teachers and facility staff. CSSF X had adopted the school curriculum in their facility branch school several years before this study. Because the CSSF X teaching and facility staff had been found to be particularly proactive in working together to create a successful environment, this institution was chosen as the ideal research site to assess the changes that were necessary. It had been observed that it was often difficult for teaching and facility staff to understand each other's practices in CSSF school education programs, and thus the relationship between education and child welfare was important for the people working at CSSF as CSSF X wished to develop an environment in which the teaching and welfare staff were able to share their problems. CSSF X had adopted several strategies to improve staff communication, for example by unifying the staffroom. The operations at CSSF X have had a major impact on other CSSF centers because of their efforts to improve relationships within the facility and because of the workshops held to assist child welfare practitioners understand the school education operations at CSSF. Therefore, describing the cooperation between the education and welfare arms at CSSF X that have been cultivated over several years can provide valuable insights into the relationship between child welfare and school education in Japanese juvenile detention centers more generally.

Analytical framework
CSSF are unique facilities in Japan as they operate within enclosed environments and are largely independent of the broader society. Therefore, the CSSF could be seen to be similar to Goffman's characterization of asylums. By participant observation in facilities such as mental hospitals, Goffman focused on the social interactions in spaces isolated from the rest of society. Goffman's theory has been useful in understanding facility care where the care receivers are similarly separated from broader society. Building on this perspective, this paper first focuses on the general features of CSSF X by drawing from Goffman's characterization of the 'the total institution': Their encompassing or total character is symbolized by the barrier to social intercourse with the outside and to departure that is often built right into the physical plant. (Goffman, 1961, p. 4) Resonating with this description, CSSF's children are isolated from the society outside of the facility. Insofar as there are no physical boundaries (such as high fences or a stockade), the children are not strictly bound. In that sense, therefore, the CSSF might be considered an 'open total institution'; it does separate the children from the society, but it also provides security and relief. Goffman suggested the term 'role dispossession' (Goffman, 1961, p. 16) as the main characteristic of total institutions. For example, when people enter total institutions such as mental facilities, they are deprived of their 'role' in the society, their 'personal belongings,' and sometimes even their name, all of which totally separates them from the outside society. Similarly, in CSSF, there is an element of 'role dispossession' for newcomers. However, the meaning in this context is a little different from Goffman's original notion; for example, removing the juvenile delinquents from the community that may have been the cause of their delinquency is seen as removing them from their 'role' as a juvenile delinquent. Further, children who come from dysfunctional families are removed from their family and given a new role as a pseudo-family member in CSSF. While this family separation deprives the children of their biological roles as a 'son or daughter,' their new role gives them security and relief from violence or other dysfunctional family behavior.
However, the CSSF 'role dispossession' is not intended to completely break off connections with the people and the society outside the facility; for example, each child wears a different school uniform when attending the facility branch school. Therefore, the CSSF could be seen to be raising children by partly connecting the activities in the facility to those in the society and without fully separating the children from the outside society. In this way, CSSF X has departed from the conventional notions of the total institution as it functions as an 'open total institution,' in which the 'role dispossession' is only partially enacted. To this end, the CSSF is organized to secure children from a society that may harm them, which is in direct contrast to Goffman's total institution, in which the people are removed from the society because they are a danger to the society.
In the next section, I present the interview data and discuss how the teaching and facility staff viewed the introduction of school education into CSSF X.

Research findings
'Entry' and 'erosion' by school education From the perspective of a total institution, how can we think about the introduction of school education into CSSF? Of course, it is thought that the introduction of school education is desirable for the children to guarantee their right to education. However, school education is seen as part of the outside society in CSSF X, because school teachers especially are regarded as outsiders by the children and welfare staff. In this way, the school education might threaten the suitable order of CSSF X as a total institution. Based on the hypothesis as described above, the experiences of the facility staff who needed to handle the introduction of school education into CSSF X were examined.
Before the introduction of school education, the 'equivalent education' program had been conducted by the facility staff who lived with the children. Figure 2 shows the content of this equivalent education.
As this figure shows, the children were moved from ordinary schools to CSSF X, and, from a micro-perspective, from the classroom of an ordinary school to a dormitory to provide security and relief to these stigmatized children (Goffman, 1963) who had been either 'delinquent' or 'abused.' Before the introduction of school education, 'equivalent education' had been conducted by the facility staff under the supervision of two teachers appointed by the education board. Facility Staff H, who was involved in the equivalent education, described the situation as follows: The facility staff were in charge of English, mathematics, or science. We decided what subject we would teach from the atmosphere in the classroom that day. For example, if a child could not understand a fraction, we taught this [. . .] Actually, the atmosphere of the classroom was steadier than now because the facility staff focused on student discipline such as quietness or posture. (Facility Staff H, 5 August 2014) To quote Goffman, 'In total institutions spheres of life are desegregated, so that an inmate's conduct in one scene of activity is thrown up to him by staff as a comment and check upon his conduct in another context,' (Goffman, 1961, p. 42). As Facility Staff H's comments show, all major guidance provided by CSSF X facility staff could be characterized under the spheres of 'Life guidance,' 'Work guidance,' and 'Study guidance.' The facility staff were able to maintain discipline as they were able to instruct the children in all spheres of life both inside and outside the classroom setting.
Following the revision of the Child Welfare Act in 1997, school education was introduced to CSSF X and established at the facility branch school, which involved the movement of school teachers into the CSSF X facility. Figure 3 shows the relationship between education and child welfare before and after the introduction of school education at CSSF.
Facility Staff J described how the facility staff interpreted the introduction of school education into CSSF X: The grouping of children between the branch school and the dormitory changed. In other words, the dormitory functions as a living group and the branch school was no longer an extension of children's dormitory lives. Before the school education was introduced, class organization was based on the dormitory unit as the facility staff thought this could provide continuity. However, this was not seen as necessary when the school education was introduced. (Facility Staff J, 7 August 2014) This narrative reveals how the classes at the branch school were not an 'extension' of the children's dormitory life after the introduction of school education. Before the introduction of school education, the spheres of life in CSSF X were not separated, so the study guidance was an extension of the life guidance and work guidance. However, after the introduction of school education, the study guidance in the facility was taken out of the hands of the facility staff and was conducted by qualified teachers. As a result, the branch school classes were basically organized according to learning level rather than by dormitory unit as had previously been the case, meaning that the classroom and dormitory groups were made up of different groups.
From these narratives, it is evident that the 'discipline' in CSSF X was disturbed by the introduction of school education due to the formation of the 'branch school groups.' In this situation, the facility staff felt rejected by the school education program. Facility Staff H described this as follows: I was in doubt about why children who dropped out of school life needed to have an education equivalent to regular school education, as we had concentrated on building the children's self-reliance so as to accommodate them to studying. We recognized that study guidance was of secondary importance. (Facility Staff H, 5 August 2014) This facility staff member expressed doubts about the value of the school education for the children in this facility. Importantly, he saw the children as those 'who had dropped out of school life.' As discussed in 4.2, CSSF X had been previously characterized as a 'gentle asylum' that separated the children from society to protect them. Therefore, the facility staff saw the school education as a potentially harmful part of the broader society because of the way that it excluded children. These staff, therefore, all shared the desire to protect these children from a 'normal' school education. These 'feeling[s] of rejection by the school education' were also present in other facility staff members' narratives; one staff member commented that the 'school teachers do not seem to feel any difficulties if they can't teach the children well. Even if the class has students with problematic behaviors, it is good for them to finish the class.' From these facility staff narratives, it is evident that there was a large recognition gap between the teachers' notions of study and the facility staff's focus on building the children's self-reliance. The introduction of school education to the CSSF marked education's 'entry into welfare' and an 'erosion by education' for the CSSF. In the next section, we focus on the narratives from the school teachers.

School teachers as 'supervisory staff and inmates'
Goffman stated that a 'total institution' was constituted by a 'large managed group, conveniently called inmates, and a small supervisory staff' (Goffman, 1961, p. 18). From this perspective, what was the position of the school teachers in CSSF X? At a glance, they seemed to be a 'small supervisory staff' as they had the role of being in charge of the study guidance for the children. However, if we look back at their move from a general public school to CSSF X (Figure 3), it could be perceived that the school teachers were also 'inmates' similar to the children in the facility. The reason for this is that their new posts in CSSF X were not necessarily in accordance with their desires. What was found was that none of the school teachers interviewed had asked to be assigned to CSSF X. Some of them stated: 'At first my image of CSSF was like a prison' or 'I didn't know what kind of facility CSSF was. ' Goffman pointed out that a characteristic feature of 'inmates' in the 'total institution' was that 'they come to the institution with a "presenting culture" (to modify a psychiatric phrase) derived from a "home world"' (Goffman, 1961, p. 23). Further, Goffman stated that this 'presenting culture' brought into the facility was 'mortified' 1 (degraded and transformed) as 'inmates' were separated from the 'home world' (Goffman, 1961, p. 24). This was necessary because the 'total institution' needed them to transform their conception of self as a school teacher in their 'home world' of school education, into a conception of self as a 'school teacher' in which their 'presenting culture' of the 'habitus of [the] school teacher' (Bourdieu, 1980) was extracted from their 'home world' and resituated into the total institution of the CSSF. In other words, while their role as school teachers was expected to remain the same, their social positioning was transposed from insider to outsider as they moved from their school classroom to the CSSF classroom.
If we interpret the social standpoint of teachers in CSSF X as both 'supervisory staff and inmates,' it is evident that the 'habitus of [the] school teacher' involved mortification. School Teacher D described their first impression of CSSF X as follows: I think that the facility staff didn't trust the teachers. Probably they thought that the school classes could be accomplished if the facility staff kept watch. . . We were also dissatisfied that we had to reduce class times because work guidance or practicing for sports festivals was more important than study guidance in the CSSF. (School Teacher D, 25 July 2014) As mentioned by Facility Staff J, in the initial school education introduction phase at CSSF X, the facility staff had been responsible for class organization based on the dormitory units. Therefore, they were often involved in the educational sphere by observing the children in the branch school. The facility staff felt that the 'faithlessness' or 'dissatisfaction' expressed by teachers such as School Teacher D might have arisen as a response to a perception of the 'feelings of rejection by the school education' mentioned by Facility Staff H. Moreover, because the facility staff partly participated in the branch school as 'supervisory staff' to maintain 'discipline' (as mentioned by Facility Staff H), the role of the teachers as 'supervisory staff' may have been weakened.
Further, it is important to note that the school education at CSSF X was limited to a morning period. School Teacher D complained that this arrangement meant that they had to reduce class times; however, the facility staff felt that the study guidance was of secondary importance. Therefore, these different perspectives created a huge disparity between the two groups. Given the 'dissatisfaction' of the facility staff, the school teachers were also 'mortified' by their deprivation of 'class time,' which was part of their 'identity kit' (Goffman, 1961, p. 29). 2 On the other hand, the school teachers were 'mortified' not only by the facility staff but also by the children. School Teacher B described their relationship with the children as follows: At first, there were about ten children in the classroom but only two or three of them communicated with me. I worried whether they would open up to me. I thought I was isolated from the children. My colleague told me that we (teachers) started at a disadvantage because the children in this facility regarded us as an enemy. [. . .] It was difficult for us to be frank with the children during class. (School Teacher B, 23 July 2014) Many school teachers have said that they were seen as the enemy by the children. During the participant observations, disobedience by the children was frequently observed. For example, some children complained, saying, 'The mathematics teacher began to write the answer although I had been solving the problem! He never sees me.' Following this, other children sympathized with him and also criticized the school teachers. Given that the school teachers had cultivated certain ways of communicating with children before transferring to CSSF X, their inability to develop the same rapport with the children at CSSF X and their struggle to adapt to the new teaching contexts could be seen as a kind of 'mortification.' As described above, the teachers had an unstable role between 'supervisory staff and inmates.' This instability on the part of teachers also made the problems each child had more obvious in the classroom. Facility Staff H member described this as follows: I think there are many children who feel comfortable in the classroom because they are able to act as they like. Even if a child who has a lot of influence says whatever he likes, it's not like he will be restrained. I'm pretty sure he's going to talk about the bad things he's done in the past and thereby maintain his position. (Facility Staff H, 5 August 2014) The facility staff member believed that the children caused problems more easily in the classroom than in the dormitory. For example, during the participant observations, children were heard talking about their past deeds, having conversations about their reasons for entering the CSSF, or talking about old friends, all of which are banned topics in the facility. These conversations occurred not during work guidance or in the dormitory but only in the classroom. The reason these behavioral problems appeared only in the classroom was that the children took the evaluations by facility staff more seriously compared with the teachers. If facility staff evaluations were bad, there would be stricter control exercised over children, similar to the treatment at juvenile reformatories (which is the highest-security prison for young children in Japan). In contrast, as the evaluations from the school teachers (for example, from achievement test scores) had little effect on the children's lives in CSSF, the children were indifferent. Therefore, the use of the teachers in the class engendered certain behaviors in the children due to the teachers' 'outsider' status within the facility and their different ways of approaching the students.
If the 'habitus of [the] school teacher' was indeed mortified in CSSF X, how can any adjustment be possible? In other words, how is it possible to cope with the disparity between school education as a 'home world' and daily life in CSSF X? In the next section, we discuss the concept of 'adjustment' that the 'inmates' could employ in total institutions (Goffman, 1961, p. 65).

'Adjustment' by school teachers
According to Goffman, while 'inmates' reside in 'total institutions' in which they are completely controlled, they perform the role required by the facility to maintain their identity. Goffman referred to this as 'adjustment,' which is 'a way of managing the tension between the home world and the institutional world' (Goffman, 1961, p. 65). This 'adjustment' could also be seen as a resistance to being 'mortified' by the facility. 3 Goffman categorized this process of 'adjustment' along a range beginning with 'situational withdrawal,' 'intransigent line,' 'colonization,' and finally 'conversion.' As described in the previous section, the teachers were 'mortified' by being appointed to CSSF X. By being considered as both 'supervisory staff and inmates,' they not only lost their positions in CSSF X but needed to employ 'adjustment' to manage the tension between the school education provision and the child welfare aspects. In this section, we focus on the current feelings of the teachers several years after the introduction of school education, and discuss their attempts to resist the feeling of being 'mortified.' One veteran teacher, School Teacher C, who had taught in CSSF X since the introduction of school education, described his current feelings as follows: Although ordinary school concentrates on studying, we also understand that the CSSF has to conduct study guidance in a situation where children can study in a mentally stable environment. Based on the situation of the CSSF now, it is important for us to change and accelerate the classes and studies accordingly. (School Teacher C, 25 July 2014) Although School Teacher D complained about the need to reduce class time, School Teacher C's remarks revealed how the school teachers' understanding had gradually changed. Another school teacher remarked that 'it's impossible for me to be with the children for 20 h [as the facility staff are], so I hope that the burden of the facility staff will decrease sometime.' This narrative indicated that the teachers respected the facility staff for supporting the children all day and were trying to accept that the child welfare instruction was a necessary foundation for study in the branch schools. In the same way, when questioned about the significance of the school education in CSSF X, all teachers stated that they cooperated with facility staff. Although the teachers had less class time and were 'mortified' by the way they needed to communicate with the children, over time their experiences in CSSF X had generated more positive feelings. How should these changes be interpreted?
Goffman called this sort of 'adjustment' 'colonization' (Goffman, 1961, p. 62), where 'the sampling of the outside world provided by the establishment is taken by the inmate as the whole, and a stable, relatively contented existence is built up out of the maximum satisfactions procurable within the institution' (Goffman, 1961, p. 62). Although the teachers had several limitations, they were able to determine the significance of the school education at CSSF X on their own.
Another important issue was how they managed the tension between the school education and child welfare aspects at CSSF X. The most important source for understanding the narratives of the teachers was the annual school festival. Many events at CSSF X play an important role in the children's life guidance, which were characterized by the relationships developed when the facility staff directly instructed the children. How did the teachers feel at these events? School Teacher F explained: I think communicating with the children outside the classroom is quite important. For example, if I positively join an event like the annual school festival from the beginning to the end, the children say to me 'Teacher, we really enjoyed it!' I think we can develop relationships with the children through such experiences. (School Teacher F, 5 August 2014) The annual school festival in CSSF X could be seen as reducing the power relationships between the facility staff and the children as they need to cooperate to put on the festival. Goffman called this phenomenon 'role release' (Goffman, 1961, p. 90) for facility staff and 'institutional ceremonies' for the children (Goffman, 1961, p. 89). Goffman claimed that 'institutional ceremonies' were 'characterized by a release from the formalities and the task orientation that govern inmate-staff contacts and by a softening of the usual chain of command' (Goffman, 1961, p. 90). Because the teachers also joined the annual CSSF X event, the relationship between the teachers and the facility staff as well as the relationship between the teachers and the children represented a type of role release. As School Teacher F's remarks revealed, from the way the children expressed their appreciation to the teachers after the annual event, such events could work to change the children's initial hostile attitudes toward seeing the teachers as the 'enemy' towards seeing the teachers as friends or allies.
The second relevant factor that contributed to the defusing of the tension between the spheres of school education and child welfare was the increasing number of younger school teachers. One young facility staff, Facility Staff I, noted: Now we [facility staff] don't just talk to the teachers when we have something to ask; instead, as young staff, we try to have a relationship with the young teachers. For example, we have held drinking parties sometimes where we as young people cooperate with each other. (Facility Staff I, 7 August 2014) Nowadays, young school teachers who have had little experience in regular Japanese public schools tend to be assigned to CSSF X. As a result, these young inexperienced teachers behave in a way that enables the facility staff to develop the relationships described by Facility Staff I, because they have come to CSSF X as 'inexperienced school teachers' who had not yet formed the 'habitus of school teachers' in the public school system. As these young teachers gained teaching experience in the child welfare facility, they were able to take the skills they cultivated to public schools, thus enabling them to apply these to a form of student guidance. Goffman defined the application of adjustment practices in one 'total institution' to other similar facilities as the 'effect of immunization' 4 (Goffman, 1961, p. 65). In a similar way, CSSF X has provided training for the young teachers that have had a kind of immunization effect.
When the tension between the spheres of school education and child welfare was defused, how did facility staff interpret the significance of school education? Facility Staff K stated the following: Also within life guidance, things that happened at school were dealt with in the school. It's good for the school to guide the children from the perspective of the school's sense of values. The dormitory is a place where the children feel safe. And when the children go to school, they get a taste of the outside society. After they graduate from the CSSF, they will inevitably be exposed to society without a place of safety [. . .] Although the children often cause trouble in school, the facility staff can identify the tendencies of how the children will commit trouble. It's just like the epitome of society; so we have to let the children cause trouble to a certain extent. (Facility Staff K, 7 August 2014) As mentioned in the previous section, the children's problems were more obvious to the teachers who had come from outside to teach in CSSF X because the 'habitus of school teachers' that they had cultivated in their 'home world' of the public school had attuned them to recognize certain kinds of problems within the branch school while also potentially provoking such behavior. As Figure 3 shows, the teachers who came from outside CSSF X were able to observe the children's behavioral problems more objectively than the facility staff. From the perspective of the children themselves, the teachers from outside CSSF X seemed to bring a 'taste of the outside society.' With this in mind, is it possible to consider the problems of these children in the classroom as epitomizing the problems that they are likely to face following graduation? It is possible to consider that the school education in the CSSF is a place where the children can temporarily experience a pseudo-society separate from their life in the dormitory with the facility staff. In such a way, the institutional changes that have occurred through the revision of the Child Welfare Act in 1997 have provided a place that may encourage the manifestation of the behavioral problems in the children.
Because these children have tended to be stigmatized as 'juvenile delinquents' after graduation from the CSSF, establishing self-reliance after graduation is very difficult. When the door to this 'safe place' is opened, children who are unable to live up to the expectations of the society may experience problems again and lose their opportunity to create a new life. From this perspective, the introduction of school education into CSSF X has created a new way to prevent the children from dropping out of society after graduation. One facility staff member referenced the contrasting time-axis perspectives between facility staff and teachers to indicate the ways that the school education and child welfare could be coordinated as part of the children's transformation process: Nowadays I think that the facility staff guide the children from a past to present perspective. On the contrary, I think the teacher's guide the children from a present to a past perspective . . . It's ideal that these education and welfare perspectives merge. (Facility Staff J, 7 August 2014) This staff member further described the fundamental education and welfare differences in terms of their approaches to the children's development. For example, he explained how there were 'time gaps' between the contrasting logic of education and welfare. When problems with a child surfaced, the logic of education attempted to explore how this child could develop the capacity to survive in society, while the logic of welfare tried to explore why this child's problems arose. Therefore, as the introduction of school education to CSSF X has continued on a trial and error basis, in addition to highlighting the different roles of classroom instruction versus life guidance, new role divisions between school education and child welfare have emerged based on their different logic and perspectives toward the children's development.

Conclusion
As mentioned above, we analyzed the narratives of the school teachers and facility staff based on the hypothesis that the introduction of school education might threaten the suitable order of CSSF X as a total institution. As this article has shown, when school education was introduced to CSSF X, it functioned differently from general public schools due to multiple factors, such as the existing institutional culture at CSSF, the 'teacher habitus' of the teachers, and the different logical processes driving education and welfare. Although the teachers were initially dissatisfied, the introduction of school education exposed problems in the children that had previously been invisible from the perspective of conventional child welfare in Japan.
In this study, we clarified the following points by interpreting the relationships between the teachers and the welfare facility staff through the lens of Goffman's analysis of 'total institutions.' First, the introduction of school education into CSSF X was recognized by facility staff as school education's 'entry into welfare' as well as the 'erosion by education' of child welfare. As a result, the facility staff's role of providing study guidance in the facility was taken away by the teachers. Second, when the teachers entered the new environment of the facility branch school, they were 'mortified' with their 'habitus of school teachers.' In other words, they were unable to teach the children as in an ordinary school; therefore, the teachers found themselves in the role of both 'supervisory staff and inmates' at CSSF X. Third, several annual events at CSSF X such as the school festival functioned as 'institutional ceremonies' and nullified the existing power relationships. Through the joint efforts and cooperation of the teachers, facility staff, and students, these events defused the tensions between the spheres of school education and child welfare. Following this, as the classroom space at CSSF X was transformed into a place where the children were able to temporarily experience a pseudo-society apart from a dormitory life limited only to interactions with the facility staff, the school education classes could be seen to be contributing to a smoother transition into and eventual rehabilitation in society.
The findings of this paper suggest that it is necessary to recognize the problems that emerged in the classroom as having positive significance for the children's rehabilitation process. From this viewpoint, it is important that new instruction possibilities are generated by visualizing the problems children have with school education. Certainly, while Japanese protective care for state wards (mainly conducted through child welfare facilities) is against western societal norms, this paper has shown that the introduction of school education and the initial conflicts between the school education and child welfare practices could encourage children to achieve social self-reliance.
Needless to say, this study has certain limitations. For example, CSSF X is an exceptional facility in which the teachers and welfare staff are able to openly discuss issues to defuse tensions. Therefore, the findings regarding the benefits of school education in child welfare institutions may not fully reflect the broader facility care situation in Japan. Further, as this study only focused on interactions between the teachers and the facility staffs, descriptions of the interactions between the teachers and the children and between the facility staff and the children were minimal. Future research will seek to clarify how the experiences of the children in this 'pseudo-society' can be analyzed from more diverse perspectives.

Notes on contributor
Shunsuke Takada holds a Master of Anthropology of education. He is currently a PhD student at Human Science, Osaka University, Osaka, Japan. His research deals with education and child welfare.