Psychoanalytic publications often contain the patient's discourse in the form of transcripts in the context of a clinical vignette. These transcripts are productions that result from a set of operations and technologies put at play by the psychoanalytic author, in the process of transmission and dissemination of the psychoanalytic discipline. The focus of our study was to investigate these transcripts and to determine first, if their narrative structure was affected by four factors (diagnosis, psychoanalytic school, gender and source of the transcript), and second, to articulate what type of self was promoted by such narratives. For this purpose, 93 clinical vignettes with transcripts, published in a recognized psychoanalytic journal, were analyzed and the effects of those factors upon the narrative structure of these transcripts, studied. Anova's results showed that the produced transcripts were affected in their narrative structure by the studied factors. At the same time, the studied factors tended to promote certain forms of selfhood over others through the transcripts.
Psychoanalytic publications consist, among other things, of topics such as the development of new concepts, meta-psychological issues, new clinical paradigms, problems around diagnosis, development and technical innovations as well as current debates around new proposals. If we don't just focus on the topics that make up these clinical publications, but also attend to their particular elements, we'll find what is known as clinical vignettes. The most important component of a clinical vignette is the transcription of a patient's discourse. In psychoanalytic publications, the author offers the reader access to the patient's discourse, either to illustrate an ongoing treatment, to exemplify a problem, to show a clinical element in the middle of an abstract development or because it is a document that presents a clinical case with certain characteristics that would allow the development of a new paradigm. With regards to clinical cases, it was Freud (1905/1962; Breuer & Freud, 1895/1957) himself who, very early on, became aware that clinical cases have the structure of short fictional stories, short narratives, and thereof his concern of not being taken seriously by the scientific community. Increasingly, clinical data in the psychoanalytic discipline consists of small vignettes rather than of full-length accounts of a case (Michels, 2000). In both cases, however, vignettes and full-length clinical cases, the speech of a patient in the form of transcripts and the construction of a story constitute the primary elements (Ferrari, 2012; Lewis, 2001; Nasio, 2013; Michels, 2000; Wyman & Rittenberg, 1992). Even though transcripts in discourse analysis are referred to the practice of recorded speech, psychoanalytic transcripts encompass different sources: notes, memory of the analyst, and digital recordings. The common practice is to build transcripts from memory and notes and not so much from recording devices. Nevertheless, we have chosen to use a broad definition of transcripts to include these sources for the transcripts in psychoanalysis.
* Correspondence concerning this article should be addressed to Dr Felipe Muller. Email:
There are two characteristics of these psychoanalytic productions that have to be emphasized. First, as just said, they are a production, in the sense that they constitute a part of a diffusion apparatus. Journal publications are part of all disciplines and are at the service of the promotion and diffusion of any discipline (Danziger, 1985, 1997). As stated above, among these productions, we are concerned with one specific element that is sometimes present in these publications: the transcripts. Secondly, in the making of these productions, there are technologies involved (Daston & Galison, 2007). In this case, by technologies we mean a series of operations done and employed by the psychoanalytic authors over the psychoanalytic experience itself, and at a more precise level, the analysand's discourse. The technologies include the registration of that discourse, via recordings, notes taking, memory recall, which also gives place to the operations of editing, of selecting and eliminating parts of that discourse, as well as the working towards coherence and meaning (Michels, 2000; Spence, 1982). In these operations, the author works over the patients' discourse according to his or her goal in one particular paper, the scientific dilemmas at the moment, the social/institutional demands, etc. In that sense, despite being presented as a verbatim discourse, transcripts are never neutral nor objective (Bucholtz, 2000). In other words, many factors are involved in these operations and are working in an unconscious way.
In this work, we are interested in one of the characteristics of these productions: the narrative structure of the transcripts that compose the vignettes, which are the product of an indefinite number of operations made upon a patient's discourse. At the same time, we are interested in attending to what type of self is promoted by these transcripts and to articulate them. Narrative approaches of the self have established these connections. Language, and narrative structure, went from having only a referential role, to having a more constitutive role (Gergen, 1991; Muller, 2016; Shotter, 1993; Taylor, 1991), that can promote certain forms of selfhood over others. We don't intend to define a specific type of self, such as the romantic self or the modern self (Gergen, 1991), the empty self (Cushman, 1990), the punctual self (Locke, see Taylor, 1989), the monological self or the dialogical self (Hermans, 2001; Muller, 2016; Taylor, 1991) or so many other forms of proposed selves. We intend to articulate and point out what forms of selfhood are promoted through these transcripts when we analyze their narrative structure. So, these transcripts, being the production of a group of psychoanalytic authors, contain ways of talking that are structured around some narrative characteristics which tend to be more predominant than others, and that are affected by different factors. At the same time, these predominant narrative characteristics tend to promote a certain implicit conception of self or selfhood, which we intend to articulate and point out.
Whether the psychoanalytic publication is one where we find a small clinical vignette or a full-length clinical case, the psychoanalytic author offers the reader the discourse or narrative of a patient, which must be defined as captured and therefore objectified, but that is, as said above, mainly produced. These transcripts of the patient's narrative offer researchers an opportunity to perform narrative or discourse analysis, as we are able to examine its elements and structure, as well as the different aspects of its content, and carry out studies considering the impact that the factors of interest may have on those transcripts.
The starting point of psychoanalysis is the patient's discourse and the work done over that discourse. First, it did so in a way where its cathartic function prevailed, and then, later, the focus was on the repressed unconscious expressed through discourse itself (Breuer & Freud, 1895/1957). From catharsis to free association, Freud sophisticated a technique that operated on the discourse of his patients, whether through their slips of the tongue, unintended mistakes and other failed acts (Freud, 1901/1960), the decomposition of dreams into different elements
in order to analyze them (Freud, 1912/1959a), or the discourse of the analysand that was formed from his associations (Freud, 1913/1959b).
From that starting point established by Freud, language began to have another place in psychoanalysis since the inclusion of structural linguistics and anthropology, as was the case of Jacques Lacan's work around 1950, or the more recent hermeneutical proposals developed by authors such as Donald Spence (1982) or Roy Schafer (1983) since the beginning of 1980. It is from these proposals that redefine and restate psychoanalysis in narrative terms that narratives gradually, and with increasing frequency, started to be an object of study and a way of doing studies in psychology. Initially, in the cognitive psychology field (Bruner, 1990), in its social realm (see Gergen, 1991), in developmental psychology (Fivush, 1994; Stern, 2000), and even in psychotherapeutic proposals geared towards narrative intervention (Madigan, 2019; White, 2007).
One of the ways in which research in the field of psychoanalysis has evolved is precisely by making use of patients' narratives. The analysis of their discourse in narrative form is a fertile ground for producing research both in psychoanalysis and psychotherapy now (for a review of the latter, see Avdi & Georgaca, 2007). Along these lines, Gergen and Kaye (1992) turn to the creation of meanings through dialogical acts in the therapeutic couple. On the other hand, Schneider's (2013) research explores the psychoanalytic process in pragmatic terms, and at the level of speech acts and role negotiation. But the narrative approach itself has also benefited from the use of psychoanalysis and the integration of psychoanalytic postulates to studies with qualitative methodologies (for example, see Pardo & Buscaglia, 2017; Thomas, 2007; for an introduction, see Midgley, 2006). Research on discursive analysis has also employed psychoanalytic concepts to deepen the reflective turn in social psychology (Parker, 1994).
Narrative analysis makes it possible to measure and produce values by attributing numbers to different aspects of a narrative. Measurements in narratives allow researchers to quantify their qualitative aspects (Habermas & Döll-Hentschker, 2017) and, from there, carry out content or structure analysis. If psychopathology is a factor of interest, Habermas and Döll Hentschker argue that there are five aspects of narratives that are the most commonly affected by it. The first four aspects - narrativity, agency, direct evaluations and reflective causal evaluations, correspond to morphological-syntactic aspects of narratives; the fifth is a semantic-pragmatic aspect, the listener orientation. Our study is oriented to evaluate several of these narrative aspects considering, at the same time, psychoanalytic psychopathology and other factors.
The narrative approach to psychopathology has developed in different ways. In some cases, the relationships between narratives and psychopathological diagnosis have been studied, focusing on the characteristics that make up the narratives of patients in the context of a certain treatment. In other cases, emphasis has been placed on the structure of the narrative developed between psychotherapist and patient or on the relationship between different forms of pathologies and narratives (see Gonçalves et al., 2002), taking into account, in some of these cases, narrative coherence as a measure , or the effectiveness or dysfunctionality of the narrative (Dimaggio & Semerari, 2001). Other studies have addressed specific clinical structures and their relationship with narratives, when considering schizophrenia (Lysaker et al., 2003) or when addressing Obsessive-Compulsive Disorder (OCD) (Mulhall et al., 2019). These works address the assumption that the discourse of the different diagnosis would be reflected in differences in the narrative structure of each one of them, and in most of these works it is intended to establish some forms of relationship between the narrative and the type of treatment needed.
Our interest differs from these last works. Even though we are interested in diagnosis, we don't intend to make a contribution on what the narrative characteristics of each diagnostic category in psychoanalysis are, which would imply a different approach. Our approach simply allows to answer the following questions: what is the effect of the factor diagnosis over the operation done by the psychoanalytic author as reflected in the analysand transcripts? And also, what is the type of self (more affective, more rational, more action oriented, etc.) promoted to the interior of its community by psychoanalyst authors when they publish a vignette that contains the analysand's discourse that is associated with a diagnostic category?
In a previous study (Muller & Bermejo, 2020) we have investigated the narrative structure of dreams included in psychoanalytic publications, studying the impact of four factors over it: the patient's diagnosis, the author's psychoanalytic school of reference, the gender of both the patient and the psychoanalyst, and the age of the patient. The study was carried out to determine how all of these factors affected the narrative structure of the reported dreams. We addressed, as we do in this study, the effect of these factors on narrative units (narrative tellings, contextualizing statements, affective-evaluative remarks and comments addressed to the analyst) and non-narrative units (such as meta-memory statements). Regarding the factor diagnosis, the results showed that the dreams of patients diagnosed as narcissistic contained more evaluative remarks, while the dreams of those diagnosed as borderline reflected a higher proportion of comments directed to the analyst. The psychoanalytic school of the author was an important factor as well, since there were more narrative tellings of non-personal facts of thought when the analyst was identified as Freudian and more affective components when he / she was identified as Lacanian. Finally, the gender of the analyst seems to be another important factor that affects the narrative structure of dreams, since we found a higher proportion of narrative tellings of personal facts, but of an affective nature, when the analyst was male.
The present work focuses not in dreams, but in discourse transcripts of analysands in psychoanalytic publications, and emphasizes the production done by the use of technologies and the operations of the psychoanalytic authors, more than the psychic production itself, as we did in the dream study. We propose to study the narrative structure of the transcripts, considering intentional states, types of facts, evaluative remarks, affective remarks, among other narrative and non-narrative elements. Specifically, when studying the transcriptions, we attend to four factors: diagnosis, psychoanalytic school, gender and source. We understand that each of one of these factors has an effect on the produced operations done by the psychoanalytic author over the analysand discourse, and as a consequence, promote certain types of selves.
On the one hand, we study how the narrative structure of this discourse is affected by the diagnosis. The study on the narrative of dreams (Muller & Bermejo, 2020) showed that diagnosis is a factor that has an effect on the narrative structure of dreams. As in that work, we consider the categories most commonly used by psychoanalysts today in Argentina: neurosis, borderline, narcissism, and psychosis. We understand that there are some discrepancies with diagnosis in the psychoanalytic community (Fink, 1997). Different authors could be using different variables over the same analysand for diagnostic purposes and although this is relevant for our considerations, our goal at this point is not to solve this issue, only to point out that among a certain group of psychoanalysts, those using certain diagnostic labels may tend to produce object-transcripts with certain narrative aspects that convey a certain type of self or selfhood.
As previously mentioned, we are also interested in studying whether the author's psychoanalytic school has an impact on the narrative structure of the transcripts. The dream
study showed that the narrative structure of transcribed dreams was affected by this factor (Muller & Bermejo, 2020). Taking into account the psychoanalytic schools that have the most presence at the local level, we consider four main schools: a) classical or Freudian; b) the English school, which includes authors such as Melanie Klein but also more contemporary authors such as Donald Winnicott and Wilfred Bion; c) the Lacanian school, whose reference is the French psychoanalyst Jacques Lacan; and d) the Contemporary French School, whose referents are authors such as Andre Green and Didier Anzieu (regarding the latter category, see Lanza Castelli, 2018). In this regard, it was precisely the psychoanalyst André Green (2000) who pointed out the usefulness of attending to the references in a publication when wondering about the way of thinking of its author. We are interested in studying whether, for example, a greater number of affective or evaluative components are observed in the narrative structure of the transcripts produced by authors from one school or another. It will be interesting to observe this in the Lacanian school, for example, to which it is generally attributed not to consider the affects in its clinic (see Green, 2005 and Soler, 2016).
The third factor that we consider important to study is the gender, both of the patient and the analyst, and the impact that this may have on the production of the narrative material found in the transcripts. In the study on dreams (Muller & Bermejo, 2020) we saw the incidence that the gender of the analyst / author had on the proportion of narrative units. In the present work, we are also interested in studying whether the produced narrative in the form of transcripts can vary according to the gender of the analyst and, at the same time, whether there are more dominant narrative elements for patients of different genders. Does the produced narrative of female patients convey, as it is more generally assumed in stereotypical thinking, more affective statements than those of male patients? If the author is a male analyst, do the produced transcripts tend to show a narrative where the agent is the patient, or where the agent is someone else?
A potential problem for our study was the source of the transcripts. Since Freud's (1912/1958) classic Recommendations to Physicians Practising Psycho-Analysis, analysts should not take notes during the sessions, with the exception of some dates and names. The reason is that the note-taking process operates as an obstacle to listening. However, the source of the transcripts does appear in a variety of ways, some being products of recorded sessions, others, of notes taken during the session, and others, of a construction based on the memory of the analyst. The source then can be an important factor, if there turn out to be important differences between transcripts which source is a transcription of a recorded session and those which source is the memory of the analyst/author.
For the study of these four factors, we focus on their influence on the differences of proportions of the different narrative and non-narrative units (dependent variables) that make up each transcript. We consider narrative units such as narrative tellings -which include facts, but also predominant intentional states-, contextualizing statements, affective-evaluative remarks and comments addressed to the analyst, as well as non-narratives units, such as meta-memory and meta-narrative statements. Thus, the objective is to study in what way the narrative structure of the produced transcripts present in clinical vignettes varies according to diagnosis, reference school of the analyst who transcribes the discourse of the patient, the gender of the analyst and the patient, and the source of the transcripts.
The material used in this study is composed of 93 clinical cases. These were all collected from articles published in the Revista de Psicoanálisis (Journal of Psychoanalysis) of the Asociación Psicoanalítica Argentina (Argentine Psychoanalytic Association) over a period of more than two decades (from 1995 to 2018). The published articles from which we obtained the transcripts of the patient's discourse had to meet two requirements: A) contain a transcript of the patient's speech of at least ten narrative and / or non-narrative units; B.1) mention diagnosis, contain bibliographic references (to identify the school), specify the gender of the patient and the source used in the composition of the transcript; or B.2) in the event that the data -for any of the items required in B.1- were not available in the publication, have the possibility of contacting the authors of the articles to consult them. In these cases, we used a brief survey with four questions, related to the factors already mentioned, and which is explained in the next section. In some cases, transcripts that did not present specific information about some factor, and where it was not possible to specify it in the interview with the author, were preserved anyway because they presented information on other factors of the study.
As already mentioned, the sample consisted of 93 cases. On the one hand, for the ‘Diagnosis’ variable, the distribution was as follows: Neurosis [26, which decomposes into Hysteria (11), Obsessive Neurosis (10), Phobia (5)], Narcissism (22), Borderline (24), Psychosis (6). There was a total of fifteen cases that did not present a diagnosis and where diagnosis was not obtained from the interview with the author. The six cases of Psychosis were discarded from the final analysis due to their low frequency. For a cross tabulation between Diagnosis and School, see Table 1.
For the School variable, the frequencies were the following: Freudian School (11), Lacanian School (13), English School (33), Contemporary French School (30), Others (for example, Kohut) (4) (see Table 1). In two cases the school was not identified.
| Diagnosis | |||||
|---|---|---|---|---|---|
| School | Neurosis | ||||
When taking the gender of the analyst into consideration, the sample consisted of a total of 48 female analysts and 45 male analysts. On the gender side of the patient, the sample consisted of 43 females and 50 males (for cross tabulation of gender, see Table 2).
| Analyst Gender | |||
|---|---|---|---|
| Patient Gender | Female | Male | |
Regarding the Source of the transcription: Transcription of recording (8), Notes (12), Reconstruction based on memory (20), Notes plus Reconstruction based on memory (15). In 38 cases, in which the author could not be interviewed, no information on the source was obtained. For the narrative analysis, a grid developed by Hirst and Manier (1996), and that was already applied in previous studies (Muller et al., 2016, 2018), was updated based on our current interest, and do not derive from the transcripts. The categories of this grid are exhaustive, and have similarities with the type of narrative analysis done by other researchers (e.g., Habermas & Döll-Hentschker, 2017). Following Hirst and Manier (1996), two main components were considered: Narrative and Non-Narrative Units. The first are divided into Narrative Tellings (which are divided into Facts -in this case, they can be Action, Thought or Affective-, and Intentional, and at the same time both can be Personal, Non-Personal or Impersonal), Contextualizing Statements, Affective-Evaluative Remarks (which in turn can be Affective or Evaluative, Positive, Neutral or Negative, and at the same time Personal, Non-Personal or Impersonal) and Comments addressed to the analyst. The latter can be Meta-memory or Meta-narrative Statements (See Appendix A).
Articles published in the Revista de Psicoanálisis were pre-selected, in accordance to the criteria explained in the preceding section. In cases where the author of the publication needed to be interviewed, he / she was contacted to arrange a brief telephone interview about his / her publication, in which four questions were asked: 1) if he / she remembered the patient whose transcription we found in the publication; 2) if so, the diagnosis of this patient (Hysteria, Obsessive Neurosis, Phobia, Borderline, Narcissism and Psychosis); 3) the source of the transcript (Recording, Notes taken during the session, Reconstruction based on the post-session memory, Other source); 4) his / her main reference authors (Freud, Lacan, Klein, Winnicott, Green, Others).
The narrative analysis on the selected clinical vignettes explained in the preceding section was only performed on the transcripts of the patient's discourse. In all cases, the literal production of the patient, as presented in the selected publications, was analyzed. The analysis was carried out by two assistant researchers trained in our narrative scheme analysis described above. They proceeded to carry out the analysis individually. Following the definition of narrative tellings, that requires that a central topic or a main theme has to be identified, the research assistants proceeded first by identifying these main themes or central topics to organize the narrative analysis around them.
In a second moment, agreements and differences in the classification of each unit were assessed together. There was an initial agreement in 90% of the units between the two raters. Differences were discussed in search of a resolution. Finally, there was an agreement in 98.6% of the units.
In cases where there was no agreement, the units in question were eliminated from the final analysis.
First, after the narrative analysis, the proportion of each dependent variable over the total units was established. Next, the goal of the analysis was to evaluate the differences in proportions of narrative and non-narrative units depending on the factors described. To do this, the assumptions were verified to be able to carry out Analysis of Variance (Anova's). In the Results section, we focused especially on homogeneity of the variances, through the Levene test. In cases of compliance, the Anova results are reported (with the Bonferroni post-hoc test in case of factors with more than two levels). In cases of non-compliance with the assumption of homogeneity of the variances, Welch's Anova's were performed, with the Games-Howell post-hoc test, if necessary. One-way Anova's were carried out, since they avoid the substantial loss of cases. In some cases, significant differences in more general dependent variables that encompass more specific dependent variables are not reported, because the variation of that variable is fully or partially explained by the latter and the differences are between the same groups. For example, the main effect for Narrative Tellings for the factor Diagnosis (more proportion for neurotics than for borderlines) is not reported. Instead, results for Narrative Tellings of Facts are reported. The reason is that there are no significant differences for the variable Intentional Narrative Tellings and differences in Narrative Tellings are explained by differences in Narrative Tellings of Facts.
A quick analysis allows us to discard the factor source, since less than 9% of the sample referred having used some form of recording as their source for the transcripts, while its vast majority was composed of transcripts based on note taking and memory, following Freud's advice. The results section is then organized by the analysis of the three remaining factors, diagnosis, psychoanalytic school and gender, in that same order.
There were significant differences in four dependent variables for this factor. First, for Narrative Tellings of Facts, F(2,69)= .82, p= 027, np2= 100. Through a post-hoc Bonferroni test for multiple comparisons was shown that those diagnosed with Neurosis (M=.75, SD=.13) produced a significantly higher proportion of these units than those with a Borderline diagnosis (M=.64, SD= 17), p=.023.
Second, for Narrative Tellings of Personal Action Facts, F(2,69)=3.14, p=.049, np2=.083. The Bonferroni test showed that neurotics (M=.29, SD=.13) produced a significantly higher proportion of these units than narcissists (M=.21, SD=.09), p=.044.
Third, for Positive Impersonal Evaluative Remarks, Levene's test was significant, F(2,69)=20.91, p<.001. Therefore, a Welch's Anova was carried out, F(2,33.39)=5.90, p=.006. The Games-Howell post-hoc test showed that those with a Borderline diagnosis (M=.02, SD=.03) used a significantly higher proportion of these units than those with a Neurosis diagnosis (M=.00, SD=.00), p= 032.
Finally, for Negative Impersonal Evaluative Remarks, Levene's test was also significant, F(2,69)=11.24, p<.001. The Welch's Anova showed significant differences, F(2,34.96)=3.65, p=.036. However, the Games-Howell post-hoc test showed that the difference in the production
of these narrative units between those with a Narcissism diagnosis (M=.02, SD=.03) and those with a Neurosis diagnosis (M=.00, SD= 01) did not reach statistical significance, p=.085.
As a complementary analysis to that reported in the previous section, we will report here on an analysis in which only the three groups that made up the category “Neurotics” in the preceding analysis were compared with each other: Phobia, Obsessive Neurosis and Hysteria. A significant difference was found in only one dependent variable.
For Personal Evaluative Remarks, F(2,23)=3.65, p=.042, np2=.241. The Bonferroni post hoc test showed that those with a diagnosis of Phobia (M=.05, SD=.03) produced a significantly higher proportion of these narrative units than those with a diagnosis of Obsessive Neurosis (M=.02, SD=.02), p= 039.
For this factor, there were significant differences in three dependent variables. First, for Narrative Tellings of Action Facts, F(4,86)=3.77, p=.007, np2= 149. The Bonferroni test showed that also when the analyst was from the Lacanian school (M=.61, SD=.16) the narratives showed a significantly higher proportion of these units than when the analyst was from the Contemporary French school (M=.43, SD =.14), p=.011.
Second, for Narrative Tellings of Personal Action Facts, F(4,86)=4.57, p=.002, np2=.175. The Bonferroni test showed that when the analyst was from the Lacanian school (M=.37, SD=.10) the transcripts showed a significantly higher proportion of these units than when the analyst was from the Contemporary French school (M=.23, SD=.11), from the English school (M=.23, SD=.15) or from Other Schools (M=.12, SD=.05), p=.014, p=.013 and p=.006, respectively.
Finally, for Negative Evaluative Remarks, Levene's test was significant, F(4,86)=5.93, p<.001. Therefore, a Welch's Anova was carried out, F(4,16.19)=3.26, p=.039. Through the Games-Howell post-hoc test it was evidenced that when the analyst was from the Contemporary French school (M=.08, SD=.08) the narratives showed a significantly higher proportion of these units than when the analyst was from the Lacanian school (M=.03, SD=.04) or from the English school (M=.03, SD=.04), p=.040 and p= .011, respectively.
For this factor, there were significant differences in two dependent variables. For Narrative Tellings of Impersonal Action Facts, Levene's test was significant, F(1,91)=9.73, p=.002. Therefore, a Welch's Anova was carried out, F(1,78.63)=6.63, p=.012. Male patients (M=.05, SD=.05) used a significantly higher proportion of these units than female patients (M=.03, SD=.03).
For Personal Affective Remarks, Levene's test was also significant, F(1,91)=7.85, p=.006. Therefore, a Welch's Anova was performed, F(1,70.26)=4.21, p=.044. In this case, female patients (M=.03, SD=.05) used a significantly higher proportion of these units than those of male gender (M=.01, SD=.03).
There were significant differences in two dependent variables for this factor. For Narrative Tellings of Non-Personal Facts, F(1,91)=6.41, p=.013, np2=.066. When the analysts were
female (M=.26, SD=.14), the patients produced a significantly higher proportion of these units than when the analysts were male (M=.19, SD=.12).
For Negative Personal Affective Remarks, Levene's test was significant, F(1,91)=18.80, p<.001. Therefore, a Welch's Anova was carried out, F(1,52.36)=4.55, p=.038. When the analysts were male (M=.02, SD=.05), the patients produced a significantly higher proportion of these units than when they were female (M=.01, SD=.02).
One important finding of the present work is that psychoanalytic transcripts have some specific characteristics. First, psychoanalytic transcripts are, as our results have shown, rarely based on audio or video records as transcripts typically do. They are mostly based on memory and notes and count as transcripts because readers understand that what they are reading was something the patient said literally. So, psychoanalytic transcripts do not adjust to the accepted definition of transcripts but conformed a specific type of transcript that should be considered in itself, just as the psychoanalytic clinical case (Nasio, 2013). In that sense, psychoanalytic transcripts do not follow any international rules for its conformation but some sort of implicit rule within the discipline that should be addressed in future studies. Second, the psychoanalytic transcripts we analyzed did not typically reflect the interaction of two speakers, but mostly the utterances of one of them (the patient), and were presented is such a way that did not allow transcripts analysis such as gaze direction and gestures. This specific conformation of the transcript may be in part due to the way the psychoanalytic experience structures its interaction and its atypical dialogue, which is between two subjects but where the subjectivity of one of them (the analyst) is controlled or left out through a series of self-directed operations (see Freud, 1912/1958). But also, in part due to the predominant schools that tend to present transcripts in such a way. Intersubjective or relational approaches like those predominant in the Anglo-Saxon psychoanalytic world promote transcripts were both subjectivities are present (e.g., Gabbard, 2008).
At a more general level, this work is an attempt to answer the following question: What type of self does psychoanalysis promote or construct through its publishing apparatus, specifically through the produced transcripts in their clinical vignettes? We intended to respond to this question by considering that there are technologies and operations employed by the psychoanalytic authors over the transcripts which we studied. The idea is that if we study these discourses in their structure, in their composition, we can find the clues or signs of the construction of a self or subject that could be articulated from their discourse.
But also, at a more specific level, our research intends to make a contribution in deciphering the discourse of hysteria, of obsessional neurosis or of the other clinical nomenclatures that traverse psychoanalysis today, but by addressing its produced and collective character rather than its individuality, as the general approach does. We intend to show, for example, that social and institutional factors affect this discourse in its production.
Firstly, if we consider the diagnosis, we find two results to highlight: the facts (and in particular the action ones) and the evaluative remarks. The transcripts of neurotic patients are conformed by a higher proportion of narrative tellings of facts than are those of patients with a diagnosis of borderline. The accounts of neurotic patients transmitted by the authors tend to be characterized by the enumeration of actions, affections or thoughts that occurred, such as, for
example, “I felt great sadness when I left that place” (fact-affect) or “He believed that I would return soon” (fact-thoughts).
On the other hand, this characteristic has a more specific mark with respect to narcissists. The difference lies in that the facts that are most prevalent in neurotics, when contrasted with narcissists, are those of personal actions. Therefore, with respect to narcissists, the transcribed narrative of neurotics appears indicating a greater involvement in their actions. For example, the neurotic narrative showed a higher proportion of narrative statements of this type: “That morning I went to the office” or “Every Sunday I worked in that café”. Neurosis as a factor tends to produce transcripts which structure reflects a higher level of personal actions, which indicates a higher level of involvement in their experience. So, this is a self that is more predisposed to actions, and more involved with them, when compared with the transcription produced for narcissistic patients.
Impersonal evaluative remarks were other indicators of differences between diagnostic groups. Specifically, the transcripts of borderline patients showed a higher proportion of positive evaluative impersonal remarks when compared to those diagnosed as neurotics. That is, comments such as “Life is wonderful” or “Parties in NY are fun” occupied more space in their narratives. In contrast, although the result was only a trend, with respect to negative impersonal evaluative remarks, narcissistic patients' transcripts showed a higher proportion of judgments such as “Today is a horrible day” than those of neurotic patients. In both cases, the transcripts produced by neurotic patients included a significantly lower proportion of impersonal evaluative remarks than other diagnosis. So, while the narcissistic subject produced by transcripts composed by the psychoanalytic authors is one whose discourse reflects low levels of personal implication, giving birth to a subject with a negative outlook about things, the borderline produced transcripts gives birth to a subject with a more positive look at the world. But both of them tend to use more impersonal evaluations, promoting a more detached and less compromised subject in both cases. Psychoanalytic authors such as Bromberg (1979) tend to point out the detachment characteristics of both borderline and narcissistic patients, but do not stress those emotional tones our discourse analysis found.
The differences in favor of neurotics in relation to narrative tellings of facts and personal action facts seem to be the other side of the coin of the findings discussed in the preceding paragraphs: that is, the neurotic subject produced by the transcripts tends to be less engaged in evaluative aspects or activities and be more involved in events than the narcissistic or borderline subjects.
Now, if we focus specifically on the transcripts of neurotic patients (hysteria, obsessive neurosis and phobia), we find an important result. Psychoanalytic authors produce transcripts of phobic patients which have a greater presence of personal evaluative remarks than those of obsessive patients. Comments such as “I'm a good person” or “In general, I'm a bit arrogant” are more characteristic of vignettes of phobic patients than of obsessive patients. A self or subject more prone towards self-evaluation, more super-ego related, is the result of those transcripts, a characteristic much more present in the produced transcripts of phobic than obsessive analysands (for an elaboration of the relationship between phobia and superego, see Gerez Ambertín, 1999).
The author's school of reference also seems to be another factor of relevance. Here we are interested in highlighting that if the author belongs to, or identifies with, the Lacanian school, his / her transcripts will be made up of a greater number of personal facts of action than those of other schools. They will form a transcript where the narrative structure highlights the
patient's actions to a greater extent than the authors of Contemporary French, English or other schools, using, for example, phrases such as “I rode my bicycle every day in the neighborhood” (fact-action) or “At that time I worked from 9 to 18 hours” (fact-action). As the neurotics, the subject produced by Lacanian authors through the transcripts tends to be more action oriented than the subject produced by authors from other psychoanalytic schools.
Perhaps this reflects the criticism that is often made to Lacan of not dealing with affects (Green, 2005), in such a way that the transcripts of Lacanian analysts reflect events to the detriment of affects and other narrative components. That said, it is the analysts of the Contemporary French school who tend to conform transcripts with negative evaluative remarks to a greater extent than the Lacanians or those of the English school, with the use of statements such as “He is a bad person” or “Life is very complicated”. In general, the Contemporary French school promotes diagnosis such as narcissism, which would relate this finding to the one discussed above where we addressed the influence of the diagnosis of narcissism. This last affirmation and the findings regarding other narrative unit, the Narrative Tellings of Personal Action Facts, which were more prevalent not only in those diagnosed as Neurotics but also in Lacanian analysts, should lead us to be cautious with this last two interpretations, because some diagnoses were more frequent for some schools (see Table 1). For example, it is not usual the use of the Borderline category among those that consider themselves as Lacanian, and in the current sample most transcripts presented by Lacanian analysts were of neurotic patients. This does not necessarily preclude us to discuss the findings in the terms used, but one alternative interpretation could be directed to the institutional weight of some schools for some diagnoses, so both a certain school and a certain diagnosis could give rise to the discussed findings.
Gender is also a factor that affects the narrative structure of the transcripts. Male patients tend to be represented by transcripts with a greater presence of narrative tellings of impersonal action facts than female patients are. For example, in our studies, narrative units such as “It was a sunny day” were more prevalent among their transcripts. On the other hand, female patients were represented as using more personal affective remarks, such as “I am very happy at this moment”. The production of both the male and female subjects seems to be aligned with the literature. A review by Wester et al. (2002) suggests that sexual differences in emotionality are small, inconsistent or limited, but susceptible to the influence of specific situational demands (context-dependent) or learned gender roles (rather than reflective of innate, basic differences in affective ability). Specifically, the evidence suggests that women are socialized to be emotional, whereas men are not. Wester et al. review shows that girls are encouraged to express emotions, with the exception of anger and contempt, through words and facial expressions. On the other hand, boys would be discouraged to express emotions, with the exceptions of anger and pride [see Chaplin and Aldao (2013) for a meta-analysis with similar results]. According to Eagly (2009), gender role beliefs imply that women are thought to be more communal (i.e., emotionally expressive, concerned with others, etc.) than men. In this sense, the transcripts produced seems to replicate this aspect of the socialization process.
On the side of the analyst's gender, female analysts reported transcripts with a greater presence of narrative tellings of non-personal facts than male analysts did, as in the expression “My friends visited me once a week”, in where the others are the agents of the actions. In turn, male analysts' transcripts were presented with a higher proportion of negative personal affective remarks, like “I feel rejected”, relatively in line with what was found in Muller and Bermejo (2020), where patients were reported to having had produced dreams with more narrative tellings of personal affective facts (“I felt overwhelmed”) when the analyst was male (a narrative unit similar, although not the same as affective remarks). The male authors ended up
producing a more “depressed” subject and the female author a subject where the others are the agents of the action.
Our results show that the produced transcripts of neurotic patients in psychoanalytic publications tend to promote an action-oriented self, not overly involved with evaluations of self and other. But also, that the evaluative dimensions are more present when the transcripts correspond to those that are considered phobic by the authors, indicating higher levels of “superego” activities in these constructed discourses, and promoting a more evaluative self. At the same time, a more detached self, prone to evaluative comments, is promoted by the transcripts when a diagnosis of borderline or narcissism is in the authors mind. Once again, a more action-oriented self is also promoted by the transcripts of the authors from the Lacanian School. From the gender point of view, female patients tend to give rise, through their transcripts, to an emotional self, while male patients are represented as being more impersonal. Finally, male authors produce a depressed self, while the transcripts of the female authors give place to an implicated self.
An Action Oriented Self, a Detached Self, an Evaluative Self, an Emotional Self, a Depressed Self and an Implicated Self are our own articulation of the different selves that are promoted in the produced transcripts when we consider the different studied factors. These factors intervene in the operations and technologies employed in the production of that element involved in many psychoanalytic publications, based on the analysand speech: the psychoanalytic transcript.
One way to consider the relevance of the study lies in the fact that it is in their publications that psychoanalysts represent the discourse of their patients. Therefore, it is in these publications with their corresponding transcripts where we will find, for example, the hysterical discourse as it is represented. But this representation process is more the result of a set of operations and productions done over the patient's discourse than the patient's discourse objectively captured, or the true discourse of a patient. At this point we contemplate the fact that in these transcripts there are two registers of subjectivity: analyst and patient. But also, we have to add a third one, that of the psychoanalytic author, who ends up producing the discourse of the patient through the operations done over his or her speech, in the process of constituting the material that will illustrate the presentation of a clinical case, or the exemplification of a theoretical or technical development. That is why our study makes a contribution to the structural conformation of the narratives that we find embodied in the form of transcriptions in psychoanalytic publications; on the one hand, the two actors of the analytic experience as source of the material, and on the other hand, the psychoanalytic author, that through a set of operations and techniques produces or creates those transcripts that transcend to the psychoanalytic community.
The authors wish to thank both Lic. Micaela Atilio and Lic. Michelle Cameron for their collaboration in this project.
| Narrative Units | |||
|---|---|---|---|
Narrative Tellings Describe states (intentional or not) or events that are linked together (causally, temporally, or spatially) and that relate to a central topic or theme | According to content | Intentional | Description of current intentional mental state “I think they're going to tell me not to go anymore” |
| Borderline |
|---|
| Psychosis |
|---|
| Total |
|---|
| Freudian | 3 | 4 | 0 | 1 | 8 |
| Lacanian | 10 | 0 | 2 | 1 | 13 |
| English | 7 | 6 | 11 | 3 | 27 |
| Contemporary French | 5 | 9 | 11 | 1 | 26 |
| Others | 1 | 3 | 0 | 0 | 4 |
| Total | 26 | 22 | 24 | 6 | 78 |
| Female | 26 | 17 | 43 |
| Male | 22 | 28 | 50 |
| Total | 48 | 45 | 93 |
| Fact |
Involves the description and narration of facts/events “In just a few weeks I lost ten kilos” |
According to subtype: Affective: “Before coming to the session, I was always anxious about whether you were going to surprise me with something unforeseen” Thought: “I was seriously thinking about going with my boyfriend” Action: “I got married to leave home” |
| According to the agent involved | Personal | Describe states or events whose agent is the narrator (active agent) “The other day I felt very bad, sad” |
| Non personal | Describe states or event whose agent is other person or group “The boss one day told me, agreeing with me, ‘they are unrecoverable’” |
| Impersonal | Describe states or events where there is no agent “The factory is growing every day” |
Affective-evaluative Remarks Provide editorial judgements or express overall emotional reactions from a current perspective to narrative tellings about past events | According to the predominant feature | Evaluative | “And everything about me seems so sick” |
| Affective | “And that excites me!” |
| According to the agent involved | Personal | “It distresses me that he is angry that I am not here.” |
| Non personal | “He had a protective presence” |
| Impersonal | “The world sucks” |
| According to valence | Positive | “But I am also happy because I am alive” |
| Negative | “It's a bit painful after I've turned my life upside down for her!” |
| Neutral | “Maybe, there's nothing wrong with it” |
| Contextualizing statements | Narrative tellings related to events or states outside the immediate spatio-temporal context of the narrative, adding “context” to the narrative tellings “Our grandfather had come from Italy in 1930” |
| Comments addressed to the analyst | “Doctor, how old are you?” |
| Non narrative units |
| Meta-narrative statements | Comments about discourse itself “I am not sure about what I am saying” | ||
| Meta-memory statements | Comments about memory and the way of remembering “This memories are fuzzy, I find hard to remember those days” | ||