Dynamic Patterns in the Voices of a Patient Diagnosed with Borderline Personality Disorder, and the Therapist throughout Long-Term Psychotherapy

Abstract This case study identified the subjective change in a patient diagnosed with Borderline Personality Disorder (BPD) by analyzing the dynamic patterns that emerged during discursive interaction with her therapist during a successful long-term psychotherapy. A qualitative analysis was conducted by applying the Model of Analysis of Discursive Positioning in Psychotherapy (MAPP), tracking voices and the personal positions of the patient and therapist in all sessions. Subsequently, dynamic patterns were identified and the hypothetical attractors were defined (i.e., the most stable patterns in the interaction of voices) using the Space State Grid (SSG) technique. Five sessions (representing the initial, middle and final stages) were selected to describe the trajectory of the patient’s subjective change. These sessions highlight the presence of different attractors and depict the intersubjective interaction that developed during the process. The results suggest a process of subjective transformation from a monological to a dialogical dimension, a change supported by therapeutic interaction based on propositional and reflective discursive aspects; a transition from a state of dissociation of the patient to a reorganization of her subjectivity. The most relevant characteristics of this process as an emergent quality of psychotherapy and its relation to the patient’s positive outcomes are discussed.


Introduction
The concept of the dialogical self in psychotherapy focuses on the nuances of self-narratives, emphasizing the scope and limitations of discourses that involve personal and interpersonal meanings (Neimeyer & Cabanillas, 2004).This concept notes that self-narratives are continuously structured and restructured by multiple voices (a metaphorical expression to refer to the aspects that form the self) that are in a dialogical exchange that constantly edits the different existing stories (Hermans & Dimaggio, 2004).Narrated stories emerge every time subjective positions are endowed with these voices, which can engage in dialogical interactions (Hermans, 2001) that give them dynamism.In psychotherapy, the self can move between two poles, one of which is monological and fragmented, and establishes the leading role of a single voice or multiple disconnected voices.The other is dialogical and multi-voiced, where the reciprocal influence of the voices that stimulate exploration and the interweaving of ideas, opinions, and feelings in the search for new possibilities is manifested (Anderson, 2012;Stiles, 1997).From a dialogical perspective, the construction of meaning is a relational product that arises from the interaction of the self with other selves, or parts of one's own self.The monological perspective, one the other hand, refers to a condition in which no responses are conceived that allow the initial state to change, resulting in the reduction of its multiplicity, or the rigid dominance of one of its parts (Gonçalves & Guilfoyle, 2006).
The transition between monologism and dialogism is marked by the reduced dissociation among voices and their engagement in adaptive and flexible interplay, which translates into the exploration, recognition, and their eventual coordination.This leads to a process in which positions can be expressed in a balanced manner, or to the strengthening of the agency that coordinates the different positions (Georgaca, 2001).

Change in psychotherapy from a dialogical approach
Evidence of clinical change can be observed through more fruitful dialogues among the different voices, decreased disorganization and dissociation (Avdi & Georgaca, 2009), and the emergence of meta-positions that oversee and integrate other positions in the self (Georgaca & Avdi, 2019).In psychotherapy, a process of integration occurs that configures a new state from elements that were already present in the patients' experiences, despite sometimes being isolated, dissociated, or denied.This means that problematic voices are excluded from the remaining personal experiences, and that there is a need to gradually build bridges of meaning that can be assimilated by the community of habitual voices of the self (Stiles, 1999;Stiles et al., 2006).This process involves a flexibilization of the patients' repertoire of positions (Hermans, 1997), a reorganization of subjectivity that begins with the recognition of the self 's relegated voices and moves toward an integration of the self 's internal dialogues, consolidating agency, authorship, and protagonism in the face of change, and expanding one's experience (Martínez et al., 2014;Martínez & Tomicic, 2019).The reorganization of subjectivity is conceived as a change in the predominance of voices and positions that forms the therapeutic dialogue, emerging as a set of reflective voices that allows the monologic voices that at first were unrecognized, dissociated, and assumed a unique version of reality, to also be present in the discourse.Psychotherapeutic changes are based in the construction of a dialogical space (integrating the different positions), and the development of a reflective metaposition (Hermans, 2003).These signs of reflexivity can be situated in the subjective expressions that the patients use to describe their psychological states, and those of others, and can thus address elements that were previously disconnected, denied, or not fully conscious.

Non-linearity, dynamical systems, and subjective change
From a dialogical point of view, psychological difficulties are established and maintained through a recursive process that implies narrowing the repertoire of available discourses and subjective positions.Change is a process that relies on the flexibility of the discourses, the complexity of the points of view, or a more fluid exchange among the different voices of the self (Avdi & Georgaca, 2009).Thus, it is a systemic process that generates trajectories over time, as well as interactions among the different aspects of the self that can be integrated.
Part of the research on change in psychotherapy has shown that its trajectory follows discontinuous and non-linear paths.Therefore, as the change unfolds, transitions occur, as well as periods of destabilization, and the reorganization of patterns (Bryan & Rudd, 2018;Hayes et al., 2007).Non-linearity corresponds to a disproportionality between therapeutic interventions and patient outcomes, which, in terms of the dynamical systems (DS) perspective, is explained as the capacity of small perturbations to trigger relevant changes, or for large perturbations to result in minor changes.Dyadic interactions between therapists and patients can be understood as a self-organized system with emerging properties, a dynamic organization that manifests itself in a non-linear manner over time (Salvatore & Tschacher, 2012;Schiepek et al., 2014;Tschacher & Haken, 2007).This dynamic organization can undergo critical moments that occur among the different significant moments of psychotherapy.These critical moments can occur before the establishment of novel patterns that grant new stability to the system and are also a manifestation of therapeutic change (Schiepek et al., 2014).
The self-organization of systems in psychotherapy refers to a collaborative process between patient and therapist that results in the co-creation of relational dynamics (Schiepek, 2003).It can exhibit an adaptive characteristic because it allows systems to adjust their levels of structural organization depending on environmental demands (Pincus et al., 2014).Moreover, emerging qualities can be understood as recursive and stable patterns in time that arise from the dynamic relationship of the elements of a system.Some of these patterns can be configured as attractors.Attractors are structures that exhibit structural stability, which causes the elements of a system (e.g., interaction of voices) to follow the same rules or behaviors as the attractor (Guastello & Gregson, 2011).As such, attractors (among others, fixed-point attractors) have absorption properties that stabilize the system with respect to other potential states.The system behaviors move toward these attractors, becoming increasingly predictable (Lamey et al., 2004).
By applying basic principles of the DS perspective, some studies have reported the emergence of differentiated attractors and transitions between patterns in therapeutic change processes.These allow the recognition of patterns that integrate the exploration skills of the therapist and discursive styles of the patient (Couto et al., 2016), a flexible configuration of attractors formed by the patient's innovative moments and their narrative content throughout a therapy, as well as a temporary regression to stable patterns of functioning (Ribeiro et al., 2011), the identification of a deconstructive and a constructive stage understood as a reorganization of patterns based on the patient's created meanings (Salvatore et al., 2010) and transitions in specific discursive patterns in high quality sessions (Salvatore et al., 2012).Additionally, the regulatory strategies formed by a reflective personal position (patients) and connected vocal quality (patients), as well as a proposer personal position (therapists) and connected vocal quality (therapists) can be identified, although they are more likely to occur during patients' change episodes (Tomicic et al., 2015).These studies have concluded that change throughout the course of psychotherapy can manifest as pattern formation and reorganization processes, or attractors (specifically fixed-point attractors), at different stages of psychotherapy.This is consistent with the systemic notion of emergent quality occurring without predetermined antecedents beyond the construction of meanings between the patient and therapist.
Changes in patients' subjectivity can be traced through a trajectory of voices expressing positions that are configured and reconfigured into increasingly stable patterns.Therefore, describing the change in a patient's subjectivity, in the context of his/her therapeutic relationship, can help to understand the evolution of that trajectory in a psychotherapy in which prototypically (according to the patient's clinical diagnosis), a process can be observed that begins with a state of dissociation (or vulnerability of the self) of the patient, which is integrated into the psychotherapy.This trajectory can be described during the successful psychotherapy process of a patient diagnosed with Borderline Personality Disorder (BPD) (American Psychiatric Association, 2013).Some clinical characteristics of a therapeutic process with a patient diagnosed with BPD have been described, following the path from the patient's initial dissociation to her reorganization of subjectivity (Martínez & Tomicic, 2019).Evidence has shown that these patients present an impoverished narrative repertoire, influenced by the narrowness of their autobiographical background and by reduced access to conscious feelings of agency, which are related to their childhood upbringing (Dimaggio, 2011).Their therapeutic processes account for a trajectory of subjectivity change that passes through these two poles, one initially dissociated, and the other relatively integrated, while at the same time it can manifest in different events that are triggered in a non-linear way, under the theoretical bases of the DS perspective.
A case study was designed to answer two research questions: 1) What dynamic patterns emerge in the voices and personal positions of a patient diagnosed with BPD and the therapist throughout long-term psychotherapy?; and 2) How are these dynamic patterns associated with the change process?The following hypotheses were established: a) Dynamic patterns will be found in the voices and personal positions of the patient and the therapist throughout the psychotherapy; and b) Differentiated dynamic patterns will be found in the voices and personal positions throughout the psychotherapy, depending on the progress of the psychotherapy.In this case study, a dynamic pattern refers to a specific interaction of voices that should stabilize over time.

Participants
The participants were a 31-year-old female patient diagnosed with BPD and a 53-yearold male therapist, a trained psychologist with 25 years of experience as a psychodynamic therapist.The reason for the consultation was due to severely aggressive behavior toward her 6-year-old daughter, and the referral was made by a psychiatrist in a specialized Mental Health Care Center located in the northern area of Santiago, Chile.The therapy was considered successful since the patient had an initial score of 123 points on the OQ-45-2 total scale and a final score of 51 points.According to the Chilean version of the instrument, scores under 73 points represent a functional population.Additionally, a reliable change index (RCI) above 17 was achieved, as the patient obtained a difference of 72 points between the initial and final sessions (Lambert et al., 1996;von Bergen & de la Parra, 2002).The session-by-session OQ-45-2 scores can be seen in Figure 1.
The ethical protocol for this study was approved by the Scientific Ethics Board of Universidad Diego Portales, and informed consent forms were signed by both the patient and the therapist, who allowed therapy material to be used for research purposes and related publications.In addition, the ethical process was approved by the Scientific Ethical Committee of Social Sciences, Arts and Humanities, of the Pontificia Universidad Católica de Chile, ID: 181108008.All psychotherapy sessions were video-recorded and transcribed following the guidelines proposed by Erhard Mergenthaler (Mergenthaler & Gril, 1996).
All speech turns in the 44 psychotherapy sessions were analyzed, although due to technical problems related to the recording, sessions 26, 38, and 43 were excluded.In total, 17306 speech turns were analyzed, with an average of 402 turns per session.

Identification of voices and personal positions through the model of analysis of discursive positioning in psychotherapy [MAPP, Martínez & Tomicic, 2019]
MAPP can be considered a three-level system.The first level contains the voices of an individual, which constitute his/her idiosyncratic expression of subjectivity in the discourse.The second level consists of the idiosyncratic personal positions of individuals, which groups together repertoires of voices that express or vocalize them.Finally, the third level corresponds to an abstract taxonomy that represents the typical organization of the positions identified in different psychotherapies.The repertoire of voices can be diverse and unique for each patient since it represents very specifically how his/her subjectivity is organized, while for therapists it may be less flexible since their voices express their point of view within a professional role (Martínez et al., 2014).Three general positions have been described for patients: Reflective, Dependent, and Independent; while for therapists two general personal positions have been described: Proposer and Professor.

The MAPP implementation procedure
Step 1: Voices identification.The first three therapy sessions were coded to identify recurrent voices in the discourse of the participants.A voice is a point of view expressed by means of a statement.These statements not only express verbalized contents, but also the perspective from which that content is mentioned.To do this, it is necessary to read each speech turn and codify it to answer the following question: What are the points of view expressed in what is being said?Therefore, a voice is a point of view expressed in discourse, which is made visible by means of statements.
In the following example, some associated statements describing feelings of sadness and guilt are underlined.These are part of the patient's "sad and guilty" voice: 174."P: It's that… I had been angry with my mom… I lived with her… then I got angry with her… and… and my husband disrespected my mom… then… I left the house… with… my husband… and my mom stayed in the house… and I left and everything… and she later… a week, two weeks later… ehh… happened… she killed herself… so I say that maybe she killed herself because of me… that she was angry with me… 175.T: mmm… (Nods) 176.P: I don't know… that left me like that… sad… it left me bad… and with a guilty conscience… like "why did I… why didn't I stay with her? Instead of preferring my husband?" (Session 1, P = patient, T = therapist) Next, the therapy sessions are analyzed, looking for the voices initially identified to evaluate the preliminary characterization, then refine it and improve its saturation.
Step 2: Determination of idiosyncratic personal positions.At this stage, the sets of voices in the psychotherapy sessions are categorized as idiosyncratic personal positions, either belonging to the patient or therapist.To do this, the voices of each participant are grouped into more abstract and inclusive categories.Each category is labeled according to the common feature that involves each specific set or repertoire of voices.The purpose of this is to answer the question: From what perspective does each of these voices speak?
In the following examples, expressions of the patient's "good for nothing" and "confused" voices are underlined, respectively.Each recount problems that involve others, either as the origin of the problem, or as someone to help solve it.They are part of the dependent subjective position: 291."P: So… that's… that's my question… because suddenly I come here… anyway… I came here because I wanted to… know… why I was so stupid… so dumb… so… maybe…(…) 321 P: It's like something tells me… like I can't… decide… decide for myself… I want to be me… I want to have a personality… and say: "I am like this, I am" I can't… because I don't know where I'm going… as you say…" (Session 1, P = patient, T = therapist) Step 3: Classification according to the general taxonomy of MAPP.The general taxonomy of MAPP represents the configuration of personal positions that are typically adopted by the patient and therapist in psychotherapy.This taxonomy emerged from the application of the previous two steps in adult psychotherapies, from different approaches (e.g., cognitive-behavioral, psychodynamic, etc.).Once steps 1 and 2 were completed, the different idiosyncratic personal positions were classified in the corresponding abstract categories.
Most of the voices identified in the first three sessions were present in some of the following psychotherapy sessions.Only two voices from the patient's dependent position were found for the first time in later sessions.
To ensure reliability in the application of the MAPP, a procedure similar to that suggested by Clara Hill and her team (Hill et al., 1997(Hill et al., , 2005) ) for Consensual Qualitative Research (CQR) was followed.Two independent coders worked on the first three transcribed sessions.Each codification was discussed and agreed upon, with the participation of a judge (the third author), and then categorized into different idiosyncratic personal positions.The data were cross-checked against the general MAPP taxonomy.Later, based on the categories of voices and personal positions already identified, their similarity was evaluated in each speech turn during all psychotherapy sessions.The judge assisted in the coding of instances in which the two coders had the most doubts and/or disagreements.Half of the psychotherapy sessions (sessions 1 − 22) were coded by the first author and coder 2 (a qualified psychologist trained in MAPP), while the second half (sessions 23 − 44) was coded by the first author and coder 3 (a qualified psychologist trained in MAPP).
Following the suggestions of Hill et al. (1997), andHill et al. (2005), a mapping of the processes was carried out to provide a more accurate assessment.The data analyses were arranged and organized in CAQDAS Atlas.tiversion 8.4 (Friese, 2020).According to one of the inter-coder agreement measures provided by Atlas.ti, the coders for the first half of the therapy sessions obtained a Hoslti Index of 70.3%, while those in the second half obtained a Hoslti Index of 75.7%.

Identification of dynamic patterns using space state grid [SSG]
The SSG is a technique developed by Lewis et al. (1999) and adapted by Hollenstein (2013), which consists of a graphical approach that allows the use of categorical data to quantify it according to two dimensions of space (x and y axis) that define the representation of all possible states (space states) of a given system over time.Thus, each element of categorical time series data can be represented as a dot in any of the cells, which is itself a state reached by the system at a given time.
The categorical data obtained from the voices were imported into the Grid Ware Software (Lamey et al., 2004) for analysis.In each cell of each grid, the x-axis depicted the categories of the patients' voices, while the y-axis represented the categories of the therapists' voices.The dots in the grid depict an event across different trajectories in time.Specifically, this refers to the time elapsed during which the patient and therapist remain expressing themselves from a specific voice before another speech turn occurrs, and which may continue to be present in the next event or stop so that other voices appear. 1 There was one grid per session, and each grid consisted of the interaction of each patient and therapist voice (14 × 6 grids).Each grid had this conformation because in total, 14 patient voices and six therapist voices were identified in the therapy.Thus, the interactions between the patient's and therapist's voices were represented in 84 possible states or cells.To identify hypothetical attractors, i.e., the most recurrent and stable patterns of the interaction between the patient's and therapist's voices in a given session, the winnowing procedure proposed by Lewis et al. (1999) was used, which consists of calculating a heterogeneity index to estimate at what level of interactions the system reaches a relatively steady state (see formula 1).This procedure was designed to identify empirically derived attractors and consists of determining which state is more probable than other states.It is composed of a series of iterations, starting with all occupied cells, and sequentially eliminating the cells with the lowest occupancy at each step.To identify which cells were hypothetical attractors, the heterogeneity values were quantified as a proportion of the first heterogeneity value (from the first iteration) then examined for scree, the value after the largest drop in proportions, i.e., a drop of approximately 50% or more (Hollenstein, 2013).

Heterogeneity
Expected Expected of Cells Where i is an index of the cell and j is an index of the current iteration (Hollenstein, 2013) Mean durations were used to perform the heterogeneity calculation in each session.The procedure began by ordering all the cells that had at least one event and progressed iteratively, eliminating the cell with the lowest average in each step.In each step, a heterogeneity score was calculated based on the observed and expected values for each cell.The null hypothesis was that all behavior in the state space would be equally distributed (high homogeneity) such that each cell's mean duration was the same.The heterogeneity values used the first value in which all the cells were included as a reference.A value of 1 was interpreted as all the interactions included were equally distributed.The iteration process made it possible to determine, depending on the case, when there was a significant drop in the proportion of heterogeneity, indicating that the duration of the events of the cells contained in the iteration were significantly different from the remaining cells, thereby identifying an hypothetical attractor.

Results
The different voices and personal positions identified in this therapy, as well as a brief description and characterization of each can be found in Tables 1-3.
The following example presents an interaction between the therapist speaking from an inquirer voice (belonging to his proposer position), and the patient speaking from a confused voice (belonging to her dependent position).This excerpt was taken from the first half of the first session: Below is an example of an interaction between the therapist speaking from a meta-analytical voice (proposer position), and the patient speaking from a self-dialogue voice (reflective position).It is taken from the second half of the twenty-eighth session: In all the psychotherapy sessions, hypothetical fixed-point attractors were identified, signaling that stable patterns in the interaction between the voices of the patient and the therapist were recognized.Five sessions were sufficient to describe the trajectory that began from the patient's dissociated subjective state to the reorganization of aspects of her self.This report cannot describe the hypothetical attractors of the forty-four sessions of this psychotherapy due to the format limitations, and because some sessions presented hypothetical attractors that did not indicate significant changes in the therapeutic interaction and/or in the patient's subjectivity regarding the chosen sessions.Two sessions were chosen from the beginning of the therapy (1 and 6) that showed dynamic patterns that reflected the predominance of the patient's dependent positions.Next, two intermediate sessions (28 and 30) were Patient's Personal Positions Description the integrative the patient impresses as being able to reflect on her dissociated aspects and generate understandings about the origin and dynamics of her problems; integrating content and emotion.from this position, the patient manifests a commitment and responsibility toward herself and others.the incapable the patient impresses as a person without the capacity to solve the challenges of daily life, trapped in feelings of fear, confusion, sadness and guilt; that as a whole, manifest vulnerability.someone unable to take responsibility for herself and others; while at the same time, she expresses the need to be helped and understood.the Detached the patient establishes herself as someone who leaves out affective bonds and feels comfortable without others.Moreover, she shows herself functioning according to what should be done, while also managing to push aside the painful aspects of her life that could make her weak and vulnerable.thus, from this position the patient shows agency and autonomy; sacrificing her desires, feelings and needs.
therapist's Personal Positions Description the Proposer the therapist performs verbal actions of inquiry and indication that promote a change of perspective in the patient.likewise, the therapist is installed as someone who invites the patient to reflect on herself, the relationship between different aspects of self and the situations relevant to her. the Expert the therapist establishes himself as someone who knows and has clarity of how matters are, specifically, how psychological problems occur, develop and operate.then the next day i'm so ashamed… that… that i feel dirty… i feel… ehh… bad mother… ehh… bad daughter… everything… everything there may be.' Envious girl voice the patient expresses envy for the childhood that others had, and that she did not.she laments her lack of affection and expresses ambivalence about her adulthood.Example: 'i feel envious… even of the affection of people… when they tell me… that the father or mother is worried about them… also that makes me sad… and i say, "ouu… why… i don't have that?"'fearful voice the patient expresses anxious aspects that interfere with daily situations.this voice expresses restlessness, nervousness, and the feeling of being accelerated.With this voice, she reports a decrease in her ability to take charge of herself.this voice transmits weakness, need for help and, at times even panic.Example: 'and then together with another person i get nervous… or i go to the street and i'm scared… and… it's like everything is weird.' angry voice the patient manifests resentment from past situations.it is lived from an attitude of rejection toward others, often venting aggressively to them, wishing to be left alone, and not wanting to be disturbed or hindered.Example: 'But all of a sudden i hate him… and i say, "why did he do this… why did he do that?"i never tell him directly… but i say… "and you are drinking?…why don't you take care of the house?"'Carefree voice the patient expresses resignation or minimization of the issues that affect her that she has not been able to resolve.this voice adopts an attitude of indifference or indolence toward issues that are important for the patient.Example: 'it's not that it's always in my head…but… i try not to let that get to me…i close that…i leave it at the back of my…head…i try not to remember that.' likewise, proposing to the patient, the search for an explanation, that allows her to understand or make sense of these contradictions.Example: 'are you worried about these things?Because i… at times… she, i see her genuinely worried… but also suddenly… i see her as… as if… she lives it as, as something a bit distant… right?'

Personal
Meta-analytical voice the therapist promotes a perspective that allows the patient to reflect on her actions, emotions, and the relationship between these, as well as on other individuals.it does so by indicating at possible relationships between facts, actions or processes that are not entirely manifested.Example: 'in that sense…umm…when that happened with your daughter…did something similar happen to you too? like she didn't recognize herself…to be doing that…to have beaten her.' self-revealing voice the therapist makes explicit to the patient how, what she says and does, resonates in him, and how this leads him to ask questions or offer interpretations about her.Example: '(interrupting) sure… my, my… my doubt at times… at times… yes… my doubt… is how much do you want to know about all that… how much would you like to know about all that…?' the Expert the asserting voice the therapist makes affirmative statements that seek to reaffirm or install truths in the discourse.With this voice, the therapist clarifies the patient's thoughts, avoiding relativization and/or rationalization.Example: 'that's an intense dialogue!' or 'i mean…most of this stuff is more your husband's enthusiasm than yours.' specialist´s voice the therapist acknowledges that he has expertise on the patient's psychological problems.With this voice, he screens the symptoms related to the general health of the patient, although not always, in association with her psychological problems.this voice also includes the discourses of psychoeducation and communication of information regarding issues related to the therapeutic setting.
Example: 'but it's also important to think that… that has nothing to do with you having a disease… or that you are insane… this has to do with something… in your life (…) you have a hard history… very hard.' Table 3. (Continued).

Personal positions of the patient voices of the patient
Characterization chosen because they exhibited patterns in which an interaction was identified between the therapist's proposer position and the patient's reflective position.Lastly, the final session (44) was selected because in it the patient's reflective position was consolidated.Dynamic patterns were identified in Session 1 (Figure 2).It can be observed that cell 1/1 (Inquirer/Continuity) appeared as part of the hypothetical attractors and was configured as a central aspect of the therapeutic conversation in this session.The therapist promoted reflective participation and focused on the therapeutic objectives, and the patient continued providing relevant information related to her reason for consultation.In this first session, cell 1/7 (Inquirer/Confused) was identified, highlighting that the patient was speaking with a voice that represents a disorganized subjective state.This is consistent with the first session, where an attempt was made to articulate the reason for consultation: The patient expressed her insecurities, conflicts, and fears, both concerning herself and her relationship with her relatives, her marriage, her mother's suicide, her father's alcoholism, the abuse she suffered in her childhood, and especially her episode of aggression toward her daughter.
Dynamic patterns identified in Session 6 are shown in Figure 3.This represents a session that could be considered to lack a particular attractor that acted as a unifier of the therapeutic conversation.The therapist was very active with his "expert" voices.Cell 1/7 (Inquirer/Confused) was present again, and cell 1/8 (Inquirer/Sad & Guilty) was added.Here, the "proposer" and "expert" voices of the therapist were linked with "dependent" voices of the patient, which could be part of more rigid dimensions of her personality.The appearance of these voices was framed in a session in which the patient commented on her self-esteem problems, the tricky relationship with her  daughter, and the ambivalent relationship with her mother during her childhood.From the patient's perspective, her mother was able to teach her lessons, sometimes by beating her, but sometimes being completely negligent in her care, not protecting her at key moments in her life, and of whom she remembers having an abortion in her own home facilitated by a person outside the family.In this session, the patient talked about the guilt associated with her mother's suicide, the need to heal to move forward, and the difficulties of talking about her childhood.
In the subsequent sessions, patterns similar to those described in sessions 1 and 6 were observed, as well as others in which only the therapist's inquirer voice and the patient's continuity voice were active, thus maintaining the rhythm of the therapeutic conversation and strengthening the alliance.Additionally, patterns that were a mixture of more than one monologic voice and active reflective voices were present, although none were clearly dominant (thus not defining a pattern in the strict sense).After session 25 there was a one-month break due to the summer holidays, which coincided with a worsening of the patient's condition, and after which she began to recover in the sessions that followed the return to therapy.
A change in the configuration of the patterns described above occurred in session 28 (see Figure 4).In this session, hypothetical attractors that only presented the reflective voices of the patient, and the metanalytical and inquirer voices of the therapist were active, together with the patient's self-dialogue voice, cells 1/1,2, and 3; and 3/2.In this session, the patient indicated that she was becoming aware of some behaviors that were hurting her, or of situations of carelessness toward her children that she was beginning to face, for example, smoking, making hasty decisions that she later regretted, feeling competitive with her daughter, or not worrying about her children's well-being and school activities.The patient noted that she felt a sense of self-esteem and that she was able to think things through, to realize the difference between acting in distress or anger, and from a calm state.Furthermore, she said she felt more like an adult, recognizing that there are situations that will always make her feel sorry for herself, and beginning to overcome the difficulties of the adult/ child duality.
The dynamic patterns identified in Session 30 are shown in Figure 5.A pattern was consolidated between the therapist's meta-analytical voice and the patient's self-dialogue voice, although in this case, the patient's grounded voice also appeared (cells 3/2, and 3), and between the therapist's inquirer voice and the patient's three reflective voices (cells 1/1, 2, and 3).A hypothetical attractor also formed between the therapist's asserting voices and the patient's three reflective voices (cells 5/1, 2, and 3).In this session, the patient pointed out that she had decided to stop working to be more committed to the care and support of her children.She complemented this by noting her decision to stop abusing alcohol, the same problem that her mother had, and which the patient hoped to overcome, especially considering her daughter's well-being, empathizing with the sense of anguish and hopelessness that both may have felt every time they saw their mothers under the influence of alcohol.
In this session, the patient began to reflect on how she could manifest her maternal role toward her daughter.Regarding her own mother, she confirmed that she felt that she understood her, even associating her alcohol consumption with an antidepressant effect.She felt that she was more willing, more attractive, and more adult, a feeling she confirmed when talking about getting her own house, where she would be able to live independently from her husband's family.She felt that she had made progress in her therapy: "so it's like I'm realizing as I'm leaving here, that I'm thinking, I'm analyzing myself ".After session 30, patterns like those described between sessions 6 and 28 were present, although there was an increasing presence of patterns (e.g., between sessions 37 and 41) similar, mainly combinations among cells, to those described in sessions 28 and 30.
Finally, dynamic patterns were also identified in Session 44 (Figure 6).In this final session, the hypothetical attractor formed by the therapist's inquirer voice and the three reflective voices of the patient (cells 1/1, 2, and 3), and by the therapist's meta-analytical voice and the grounded voice of the patient (cell 3/3), appeared again.It is possible that in this session, a subjective state of the patient was consolidated, established from a conscious and realistic perspective, concerning the different present and prospective responsibilities.The patient indicated a readjustment in her relationship with her husband, considering her new, more independent life, and trying to resolve the recent tensions between them, during a process in which she positioned herself symmetrically around him.She talked about the plans she had for finding a new home and the more active attitude she hoped to adopt in terms of its organization and maintenance.
Regarding her relationship with her daughter, she mentioned that she had gone from feeling that she was not her daughter, that "she was not part of my body… as if I had not had her," to being affectively and functionally committed to her, and able to express herself and have important conversations with her.The patient could distinguish between her past and her present life, saying that she understood what she suffered as a child and that now she could live differently.A conclusion was drawn about the purpose of her therapeutic work, the benefits it had brought her, the intention to continue pharmacological treatment, and at the same time, she pointed out that she needed to discuss her problems with friends who could provide her with complementary perspectives.

Discussion
The chosen sessions show the trajectory of change in subjectivity through variations in the dynamic patterns of the patient's and therapist's voices.In session 1, the interactional dynamics observed were consistent with those that could be found in any psychotherapy.Along with this, in the initial segment (sessions 1 and 6), patterns can be observed that could describe a collaborative attitude toward therapeutic work, just as an interaction where disorganized aspects of the patient's personality were displayed, which is a possible characteristic of a patient diagnosed with BPD who has had difficulties in developing her identity (Fonagy et al., 2011;Fonagy & Luyten, 2016;Richetin et al., 2017).Patterns can also be observed that involve aspects of emotional intensity that denote both sadness and guilt, which at these moments seemed to assume a leading space in the patient's experience.Despite the presence of this type of pattern in these initial sessions, there was another type that comprised self-reflective actions on the part of the patient, which allowed her to unfold her problems and try to address them actively.Dimaggio and Stiles (2007) have described problems of confusion in patients experiencing intense and contradictory feelings, which leads to disorientation and reflects incoherent inner worlds, possibly related to a lack of dialogue among different parts of the self.In this study, confusion could be recognized as one of the patient's dependent voices and was part of the relevant patterns in the first half of the psychotherapy.The decision to consider it as a voice and not the result of a lack of connection among her different voices was based on the fact that its appearance always coincided with the traumatic episodes that the patient suffered; thus, it could be recognized as the expression of an aspect of the patient's dependent position.Certainly, this may have been one of the most important problems that the therapist and patient had to face to move toward differentiation and coexistence of the different aspects of the patient's self.
Session 28 highlights a pattern formed by the interaction of the therapist's inquirer voice with the three reflective voices of the patient.From this session onwards, as shown in Figure 1, the patient's outcomes were more positive than during the rest of the therapy.Whereas the session 30, presented dynamic patterns with a configuration different from those that appeared in the previous sessions, and included a pattern formed by the therapist's meta-analytical voice, and the three reflective voices of the patient.This inclusion of the therapist's meta-analytical voice could have encouraged the integration of not only the content relevant to the patient but also of her three personal positions.It can be assumed that this was a session in which there was a high elaborative potential, and relevant issues were addressed.This is one of the patterns that could bring the most therapeutic benefit to the patient, and as an inspection of the patterns of the previous sessions suggests, it can be considered as an emergent state in the process, a property of the interaction among the therapist and the patient.
Finally, in the last session, there seemed to already be a consolidated pattern, specifically, the one comprising the therapist's inquirer voice and the three reflective voices of the patient, which enabled the association among different aspects of the self.In addition, a composite pattern of the therapist's meta-analytical voice and the patient's grounded voice had been configured, which possibly included a projection of the gains she had made in the process, and what she might face once she no longer had this psychological support.It is possible that these positive outcomes of the therapy are associated with the quality of the patterns that were formed in the process, and which sustain these changes through the particular subjective structure that was consolidated.
From the dialogical self perspective, one of the objectives of therapy is to establish historical continuity and bridges among the multiple dimensions of subjectivity, sustained by the emergence and involvement of reflective subjective states (Georgaca, 2001(Georgaca, , 2003)).The formation of reflective patterns during psychotherapy generated an experience of continuity in the patient's (and the therapist's) subjectivity.Her reflective voices allowed her to build bridges among the different voices, increasing the polyphony of the self, as the patient became more aware of the presence of her voices in different relational contexts, and was able to speak of her monologic voices from her reflective position.
To understand the differences among the different hypothetical attractors that occur in the therapeutic process, these state changes can be described as discontinuous jumps that are consolidated as all the constituent elements (i.e., the interaction of the voices) are reorganized.It is important to remember that the process of psychotherapy occurs in the form of nonlinear and discontinuous changes, as postulated in complexity theory and self-organizing systems (Schiepek et al., 2013).Salvatore et al. (2008) consider that psychotherapy does not have a predetermined or inherent construction of meaning, and dynamics of meaning emergence can be observed in the form of discrete transitions.This encourages monitoring the behavior of the therapeutic system in the stability-instability continuum, suggesting not only the future description of phase transitions but also other properties that can be linked, for example, to the slaving principle, by looking for the emergence of synchrony at some stage of psychotherapy, which is at the basis of pattern formation (Schiepek, 2003).In some studies, non-verbal synchrony between patient and therapist has been recognized, which generates greater order with time (Ramseyer & Tschacher, 2014, 2016).This type of order emergence can be compared to the formation of hypothetical attractors, which, while they do not necessarily involve synchronized behavior or enslaved processes, does require a particular form of stable interaction to be reached over time.
The present analysis emphasizes the interactional aspect of therapeutic conversation, the temporal-spatial arrangement of the encounter of the therapist's and patient's voices as the process progresses.In this context, the creation of meanings is not understood as the production of particular content, but rather is subject to a structure based on the activation of an intersubjective space that delimits the possibility of certain contents and meanings.In other words, the creation of meanings is set in the conditions of possibility provided by the different interactions of voices that could represent dimensions of the intersubjective space, which can be graphically observed in the state space displayed in the SSG.In the case of the patient, this structure can be seen to vary, for example, between the first analyzed sessions (1 and 6) and session 28 or 30, assuming that the personal positions that are interacting at key moments of those sessions allow (or at the same time, inhibit) the emergence of contents that describe how patients and therapists are understanding what is being discussed in those particular segments.The variation among dynamic patterns can be understood as a transition from one state to another, and it can be discussed whether it is a transition from a rather monological to a dialogical form, i.e., the move toward a state of openness that can come from a state in which patients have a relatively passive attitude concerning what they are dealing with in psychotherapy (Gonçalves & Guilfoyle, 2006;Gonçalves & Ribeiro, 2012).If the patient speaks from the voices of her dependent position, subjectively she may be expecting the satisfaction of a need related to someone or something external, without further involvement.Considering the rigid character of these voices, which do not regard alternatives and unfold in a predominant and pervasive way in the discourse, this could lead to a monologic state that is not helpful for the development of the treatment.If the therapist continues to speak from the same voice and/or position, the patient may remain in this monologic position, maintaining feelings of fear, confusion and guilt while waiting for a solution to be provided.The therapist's transition from one of his voices to another, either in one or both personal positions, could be a way of stimulating openness to aspects that are not sufficiently connected in the patient's subjectivity.In practical terms, this means talking about what she has not talked about before, reflecting on what she is talking about, and ultimately reducing monological interactions, allowing dialogical interactions to be achieved.
Another aspect to highlight regarding the formation of reflective patterns is the quality of the continuity voice in the patient's triad of reflective voices.This voice can be considered, as previously mentioned, as a cooperative aspect in therapy and not in itself dialogic.The study by Viou and Georgaca (2020) notes that reflective voices can have different qualities, differentiating, for example, between compassionate and critical voices, which underlines the importance of recognizing the qualities of reflective voices.In our study, the continuity voice could be considered to support the activation of the other two reflective -dialogical voices of the patient.
Through MAPP, it is possible to describe a therapeutic process based on the interactions among the voices of the patients and the therapists.It shares a similar approach to the qualitative method of analyzing multivoicedness (QUAM), having this the capacity to analyze internal and external positions, and their interactions (Kay et al., 2021), while MAPP it focusing on the internal positions of the patients and therapists.The fact that these voices can be ordered as interactive sequences of time series through the SSG allows access to a longitudinal perspective, where complete sequences of a session are explored, rather than individual aspects evaluated at a cross-sectional and isolated point in time.Although like all non-conscious processes, recognizing voices and their influence in psychotherapy can be relatively hard work, it is possible to do so in a supervised setting, whether dyadic or in a group setting.Repeating the recognition exercise in several cases can result in the construction of a repertoire with more stable personal positions, which can be situated in the psychologists' therapeutic resources.

Limitations and future research
The choice of whole-session analysis had methodological value considering the dynamic quality of the object of study: the trajectory of the change in subjectivity.As a case study, it focused on the salient aspects of a process of change over time.From a DS perspective, individual differences in change trajectories are not associated with sampling or measurement error but are part of the manifestation of change (Polman et al., 2011), referring to the heterogeneity of patients' responses to the same type of treatment, highlighting its non-ergodic character, and the need to apply methodological designs that consider idiographic and nomothetic stages.This is only one stage of this study of change in subjectivity from the dialogical point of view, but with these results, future designs can be planned that integrate and analyze a larger number of cases.One of the aspects to be explored in future studies will be to explain how the patient's monologic voices have been elaborated and integrated into the discourse of the reflective voices and acquired the protagonism they have achieved.Specifically, this refers to how reflective voices are transforming the presence of monologic voices in discourse.Another challenge will be to recognize which segments of psychotherapy to select so as not to underestimate the trajectories of change, as this study had to include all sessions to initially assess which were significant.Undoubtedly, there are other variables involved in the emergence of this type of change in psychotherapy, and it is critical to consider these potential variables in both the process and outcome.This is necessary in the search to recognize, among other things, the processes of internalizing others in the subjectivity of patients and the concrete contribution of the therapist during the transitions among the identified state changes, i.e., from initial patterns to reflective patterns.Future research could not only help in understanding the variants of this trajectory, but also the elements that facilitate its realization.

Figure 1 .
Figure 1.oQ45-2 session-by-session scoring.the cutoff score for the functional Chilean population (73 points) and the trajectory of the scores in the chosen sessions from data analysis are also shown.
the therapist performs interrogative actions, such as questions, signals, reflexes or clarifications; that have therapeutic intentions. it promotes the generation of a reflexive movement, placing the patient at the center of that reflection.Example: 'When you talk about guilt… ehh… was it because of what happened with your daughter… because of… (reformulates) or was there guilt also with something about your mom?' Confrontational voice the therapist points out aspects of the patient's speech that are absent, contradictory or inconsistent with her non-verbal expressions.

Table 1 .
Patient´s and her therapist´s voices, personal positions, and MaPP taxonomy.

Table 2 .
Description of the personal positions in psychotherapy.

Table 3 .
Characterization of the different voices belonging to each personal position.pertinentdata to the therapy; dates, events and relevant information for the progress of the session.there is not much elaboration on what is said.theinformation and data provided are usually enunciated in past or a present perfect tense and frequently account for events that occur outside of the therapeutic encounter.Example: 'i didn't drink alcohol… i didn't smoke… then i met my partner… and with him i started to do things that i didn't do… well, i tried cigarettes… i tried alcohol' self-dialogue voice the patient reflects on herself and her dissociated aspects, taking distance and perspective from a coherent and regulated emotionality.thisvoice allows her the expansion of understanding about what happens and what she is doing, integrating content and emotion.Example: 'Do you also know what is happening to me?…like… that i feel like… thinking some things… like i feel like i want to imitate her [her mother].' grounded voice the patient expresses a commitment and responsibility to herself and others, especially her children.the grounded voice is one in which the patient knows what is happening, validates and accepts her feelings, can adapt to situations, can give affection without feeling weak, and is involved in life with interest.Example: 'no, if i'm okay..' , i said… but then time passed and i said, "it's not because i'm wrong… i'm not right with myself…" then i came here [to therapy]… i'm interested… because i need… to heal.' a feeling of incapacity in her own presence and that of others.this voice is characterized by a high level of self-demand that leads the patient to continually devalue herself.Example: 'like it doesn't… it doesn't matter and then i do things… Confused voice the patient is involved in a sense of dissociation, of not knowing what she is doing or why she is doing it; all of which results in a loss of agency.she is trapped in actions and desires that she feels strange and despairing, further losing further contact with herself.Example: 'but, i still don't realize… what i have… i don't know who i am… as if i act in one way… and then somewhere else i am another person.'sad and guilty voice the patient shows constant self-reproach regarding the patient's behavior, especially concerning her mother and daughter.this voice has a tone of grief and sadness as she blames herself, because her way of being and acting harms others.Example: 'i don't know… that made me feel sad… like… sad because… that made me feel bad… and with a guilty conscience… as with guilt that "why did i have…?why didn't i stay with her [her mother]?"' Desire voice the patient shows ambivalence.the patient impresses by seeking to be looked at, by men and to seduce them, but regrets it; feeling ashamed and guilty.she often refers to herself from this voice as 'cheeky' .Example: 'no… i'm not going to do that… and then i still say "no if… what's wrong with it?if i'm having a good time, and no one else knows about it…" (laughs) like i feel that… i make advances toward men.' fear of failing voice the patient expresses fear that her daughter will live what she suffered in her childhood (e.g.violence, sexual abuse).this fear is directly related to an unrealistic fear of acting like her own mother acted with her, when she was a child.Example: 'yes… yes because… it fears me… because i think that maybe i'm going to make a mistake… or that… my daughter is… i don't want her to judge me too… like if she sees me drink… do things… that i forbid her.' the patient focuses on what she should do, disregarding the reflections on its meaning. in addition, the voice replaces and invalidates its actual feeling for 'what it should feel' Example: 'But without wanting to, it finally taught us that (…) like that i get down… like that… and when i get down… i feel bad… i say, "no, i can't be like that… weak… i have to be strong."'Disaffectionate voice With this voice, she reveals how unnatural the role of a mother is for her, and her disinterest in adopting it.this becomes especially evident when she refers to her daughter, not only expressing her carelessness and lack of interest in her, but also feelings of rejection before her affective requests.Example: 'i feel like i don't love her… or suddenly she's crying… suddenly i feel like laughing… i see her crying as well as… "ah! that is tragic![disaffected expression]" i say… things like that.'