Assessment of affective and conversational trajectories in psychotherapy with adolescents suffering from borderline personality disorder

Recently, machine learning methodologies and affective computing have become more popular in the field of psychotherapy research (AaVes-van Doorn et al., 2020; Poria et al., 2017). Methods facilitate video, audio and text processing in order to extract verbal (Goldberg et al., 2020) and nonverbal information (facial expression: Arango et al., 2019; postures: Zhang et al., 2018; paralanguage: Crangle et al., 2019). A promising application for automated streams of nonverbal information is emotion recognition (Halfon et al., 2020; Sharma & Dhall, 2021). .e drastic rise of possibilities and the eclectic use of symbolic information streams begs for anchoring theory to inform meaning making of nonverbal phenomena. Most publications study the mutual interdependence of nonverbal signals in patient and therapist (synchrony, attunement, concordance, coordination), taking a firm stand in generalizing theories (common factor theory, communication theory, information processing theory, self- organisation theory), implying that studied processes of nonverbal exchange translate to transtheoretical concepts of psychotherapy research and human interaction in general (Koole & Tschacher, 2016; Laroche et al., 2014; Salvatore et al., 2015). On a more fine-grained conceptual level, empathy, emotion regulation and quality of therapeutic (working) relationship (alliance) are proposed as correlates (Imel et al., 2014; Reich et al., 2014; Soma et al., 2020). In this dissertation, I illuminate the use of nonverbal signals from a different perspective. Being a practicing therapist myself, I anchor my thoughts in a clinical standpoint, applying theory on mechanisms of change in relational psychoanalysis (Adolescent Identity Treatment: Foelsch et al., 2014; Transference Focused .erapy: Levy et al., 2006) relevant to the psychiatric understanding of (borderline) personality disorders. By collapsing assumptions of relational psychoanalysis (Tufekcioglu & Muran, 2014) and the events paradigm (Timulak, 2010), psychotherapy is conceptualized as the mutual communication and regulation of self-states in an ongoing negotiation process that brings forth key events. More precisely, using object-relations theory (Kernberg, 1995), key events can be defined as episodes of integrative work with the dominant object relation dyads. .e instrumental involvement of nonverbal communication streams in this process is discussed (Dreyer, 2018; Mac Cormack, 1997). On the conceptual level, it is possible to find common ground between emotion regulation (Campos et al., 2011) and alliance negotiation (Eubanks et al., 2019) as they both emerge as nonverbal communication acts in moments of therapeutic work, moving towards the clinical goal of identity integration (Jung et al., 2013; Schlüter-Müller et al., 2020; Schmeck et al., 2013). I discuss how the automatic assessment of nonverbals can help recognizing these key moments. Interpersonal coordination processes have been proposed as correlates for emotion regulation and alliance negotiation. I highlight problems with the concrete adaptation of coordination assessment under the assumptions of the events paradigm and how they can be overcome by the perspective of relational approaches.