In family therapy, the way in which interaction between the therapist and the family members proceeds is a key factor in having family members continue the sessions. As the session unfolds, the therapist becomes part of the environment surrounding the family, and through the intervention of the therapist ― a newcomer ― in the existing family norms, a new therapeutic system is established jointly with the family members. Therefore, the initial session is an important stage and whether family members continue the sessions or not depends on how smoothly the therapist can become a new part of the existing group norm. This study showed how the therapist and each of the family members used an evaluative lexis and expressions to construct each social personality and interpersonal relationship of the members in a session, and conducted analysis regarding the characteristics derived from the use of evaluative language by three therapists, as well as the deviation of the effective interactions from the viewpoint of establishing interpersonal relationships. Martin developed Appraisal theory, which systematizes the linguistic evaluation of any text, via the theoretical framework of Systemic Functional Linguistics. This study adapted Martin's theory to the transcripts of three sessions of the same setting of family therapy. First, linguistic evaluation mapping was constructed, which was followed by quantitative analysis. This study focused on qualitative linguistic analysis based on the quantitative results obtained.

　　From the viewpoint of the manipulative mechanism of the linguistic interaction in constructing the interpersonal relationships of the members, discussions were conducted on (1) evaluation of “emotion” as the focal point of negotiation, (2) evaluation as psychotherapeutic verdictives combined with certain lexicogrammatical resources, such as relational process clauses and mental process clauses with the second person subject, (3) the direction of the session in which the difference of inscribed/evoked appraisal leads, (4) the evaluative reaction lexis as a factor in controlling and organizing the session, and (5) family therapy as power play.

　　The purpose of this study was to examine the association between two types of domestic power (referent-expert power, reward-coercive power) relationships (marital, father-child, motherchild) and adolescent aggression toward parents. The research participants were 247 undergraduate and graduate students (125 men; 122 women). Results showed that reward-coercive power was positively correlated, and referent-expert power was negatively correlated with aggression toward parents in adolescence. Moreover, adolescent aggression toward the father was mainly associated with adolescent evaluation of the father’s reward-coercive power over the mother. In contrast, adolescent aggression toward the mother was related to adolescent evaluation of the mother's reward-coercive power. Furthermore, one parent's evaluation of their partner's rewardcoercive power was associated with adolescent aggression toward the other parent. However, physical aggression toward the mother was associated with the mother's evaluation of the father's reward-coercive power. The result of this study implies that the influence of power relationships on aggression toward a parent differs by parental role.

　　This study investigates how adolescents in the pre-adulthood perceive their grandparents' engagement toward their socialization. Based on a semi-structured episodic interview scheme, questionnaires were conducted among fourteen male and female undergraduate students, inquiring about (1) the interviewee's favorite grandparent and that reason; (2) the meaning of grandparents' existence for the interviewee; (3) factors thought to have been inherited from grandparents; and (4) memories with grandparents. The analysis was conducted by sorting the descriptions by difference of perceived meaning of the functions of grandparents according to a temporal sequence of the grandchildren's respective life stages from verbatim records of and the interviews. From the analysis, it was observed that playing with grandparents in early childhood promotes the grandchild’s emotional exchange, and the experience of engaging in playful activities enjoyed by the grandparents promotes the grandchild's emotional expression and gender role conception. Grandchildren of school age and in puberty appear to learn sociality not only directly but also indirectly from grandparents as social references while receiving emotional support from the grandparents. Grandchildren in adolescence tend to reform their values by reevaluating the lives of their grandparents through an attempt to accept the aging and death of grandparents. The grandparents-grandchildren relationship consisting of opposite directions of aging and growing are assumed to have contributed to the socialization and emotional development of grandchildren. In addition, negative emotions toward grandparents observed in early childhood appear to directly reflect the parent-grandparent relationship, but they tend to transform into positive perceptions based on the grandchild's multilateral assessments.

　　The purpose of this study was to investigate the possible influence of mothers' childhood attachment patterns on their own parenting in adulthood. The participants of the study were mothers (n=201) who had children attending kindergarten, and their children's kindergarten teachers (n=16). The mothers were classified into three categories: avoidant, secure, and ambivalent, according to how they answered questions about their relationships with their own mothers in early childhood based on Ainsworth's attachment pattern. The data were analyzed in order to investigate the relation between mothers' attachment patterns and four factors: attitude toward child-rearing, acceptance of maternal role, marital mutuality, and their children's behavior at kindergarten.

　　The findings revealed that secure mothers had more receptive attitudes toward child-rearing, perceived their maternal role as more positive, and reported higher marital mutuality, compared with avoidant and ambivalent mothers. Avoidant mothers perceived their maternal role more negatively than did secure mothers. Children of avoidant mothers had more behavioral problems in kindergarten than children of secure mothers. The results showed that mothers' early childhood attachments to their own mothers influenced their parenting, and that avoidant and ambivalent mothers tended to have more difficulties in child-rearing than did secure mothers. It also suggests that secure attachment in early childhood is important for individuals who will become mothers and rear children themselves.

　　The measurement of maternal sensitivity needs to be accurate, even for short observation time of interaction between mother and child. The object of this research is to examine the possibility of measuring maternal sensitivity using Nursing Child Assessment Teaching Scale (NCATS) over a short period of observation. I recorded free play between mother and child (6-7 months old) using toys for 5 minutes in 65 cases. Maternal sensitivity was determined using NCATS. After approximately one year, I measured child attachment using AQS, and studied the relationship between such attachment and maternal sensitivity. I conducted a multiple regression analysis on the relationship between child attachment stability and NCATS. A significant correlation was obtained when we used sensitivity to the child's cue as an independent variable.

　　Thus, it was proved that NCATS was useful to some degree in the measurement of maternal sensitivity over a short period observation of free play between mother and child.