In this paper, I respond to the following question: What processes and mechanisms explain these limits in how educators represent parental engagement? I do this by tackling educators' representations of parents' environments and engagement. I conceptualise representation with respect to categorisation and I describe how educators categorize parents living in poverty. Cutting parents'
... [Show full abstract] environments, selecting certain aspects in the environment and representing certain zones as dense and others as sparse are the three mechanisms that I cover. In order to expand the notion of categorisation and the tension between fullness and emptiness, I present the concept of social representation from a systemic perspective and I refer to social representation dynamic as a tension between verbalism (emptiness) and ellipsis (fullness) in relation with my own theoretical work (Boulanger, 2016). Regarding my analysis of the discourse of teachers participating in a partnership program in Canada (Quebec), I situate this dynamic in school-family relationships in order to grasp teachers' representations of parents. From the 1990s onward (cf., Hoover-Dempsey & Sandler, 1995), researchers in the field of school-family-community partnership refer to many concepts-beliefs, attitudes, perceptions, etc.-that pertain to representation. In particular, educators' representations of the child and the parent are considered an important determinant of educators'-professionals acting in school and other communities' organisations and performing an educational function-practices, actions and their interactions with the parent. This is central to understanding the disagreement between educator (particularly the teacher) and parent. This is particularly for families in poverty-related context, that is, when family members face the social problem of poverty and have to deal with the associated sociocultural and symbolic phenomena, such as the negative representations of educators toward ''poor'' parents and children. But the numerous models and typologies generally fail to grasp the process and dynamic underlying educators' representations because these models and typologies tend to be associated with a cognitive-behaviorist and associationist logic (see Boulanger, 2018). It is no surprise, then, that these approaches can't help improve the