Abstract
This article presents an epistemological reflection on two models of the group: S.H. Foulkes’ matrix model and René Kaës’ less well-known group psychic apparatus model. These two models are viewed as complementary, in the sense of George Devereux’s ‘complementarist’ approach, provided that each is conceived of as constituting a ‘vertex’, as Bion used this term. After showing that the two models are neither in competition nor coextensive, we explore their complementarity. The concepts of the ‘internal group’ and ‘unconscious alliances’ proposed by Kaës provide a way to understand how communication networks emerge and are structured in the matrix. Conversely, the concept of the network of communication can also be thought of as a metaphor for the psychic spaces of the group psychic apparatus. A clinical illustration involving a group supervision session is presented in support of these proposals.
Introduction
This article is a continuation of our effort to bring together the matrix model and the various models of the group that have been introduced by psychoanalysts in France. In Vollon, Gimenez and Bonnet (2015), we explored how Foulkes’ (1964) conception of the matrix can be thought of in relation to Anzieu’s (1985; Anzieu et al., 1987) model of the group as envelope. This allowed us to draw out another approach to the network, which we came to view as a sort of neural wiring with a plastic quality (Vollon et al., 2015). However, the envelope model has been refined and extended since it was first introduced by Anzieu, especially by the psychoanalyst René Kaës. Kaës’ clinical work with Anzieu in the group setting, along with his epistemological research undertaken from 1979 to 2015, led him to construct a more complex model with major metapsychological implications: the ‘group psychic apparatus’ or ‘GPA’ (Kaës, 1976). We would like to suggest that the matrix and GPA models can be understood as two ‘complementary frames of reference’ (Devereux, 1978). In other words, we consider the models to be two specific ‘vertex’ (Bion, 1965: 91–2) that complement one another.
Kaës’ GPA model and his concept of the internal group give us a way to conceive both of what happens in the inner world of the individual—who is a nodal point of the networks of communication—and of how such networks are set up. We propose that, in the unconscious, the organizational logic of group communication networks can be at least partially understood in terms of unconscious links and alliances (Kaës, 1979). We shall begin by presenting the main characteristics of the two models and our understanding of them, and then, to illustrate our conclusions, we shall provide a clinical example from a supervision group.
The origins of the construction of the matrix and GPA models
Foulkes (1946) first presented the essentials of the matrix model in a short article entitled On Group Analysis (Foulkes, 1946). This model arose from three influences: Goldstein’s (1934) conception of neurobiology (the organism is a whole whose parts can only be grasped within its larger context); Freud’s psychoanalysis (see Foulkes, 1983: 7–9); and Mannheim’s (1943) socio-analysis, in which the study of the group provides a way to understand larger and more complex group dynamics (cf. Foulkes, 1990: 129). From the work of these authors, Foulkes draws the premises of an approach to the group as a whole. In Group Psychotherapy: The Psychoanalytic Approach, Foulkes and Anthony (1984: 258) put forth the idea that the group as matrix ‘can be thought of as a network in quite the same way as the brain is a network of fibres and cells which together form a complex unit’. These transpersonal networks ‘can be regarded as the operational basis of all mental processes in the group in the same way as the individual’s “mind” is the operational basis of all mental processes in the individual’ (Foulkes and Anthony, 1984: 258). The individual here is conceived of as a nodal point of these networks, as if suspended within them. Further, the members of a group ‘are in a state of interaction, in a common field, in interpenetration and communication. They speak now through one mouth, now through another’ (Foulkes and Anthony, 1984: 259). This is why Foulkes (1984: 66) also uses the term ‘network of communication’. Later, he would revisit his model, drawing from it a genuine psychopathology of networks: If our patients are not seen to be in need of help entirely for their own sake, but are in fact part and parcel of a whole network of interacting individuals, it follows that in certain respects they are merely the victims or scapegoats, or otherwise symptoms of changes and upsets within the intimate network of their human relationships. (Foulkes, 1990: 225)
By extension, each pathological structure would correspond to a specific network organization. Finally, Nitzgen (personal communication, 2020) pointed out that Foulkes repeatedly argued that the group ‘tends to speak and react to a theme as if it were a living entity’ (Foulkes and Anthony, 1984: 238; cf. Foulkes, 1984: 118).
On the other side of the Channel, Kaës (1976) was working on the GPA model, which he developed in his book L’appareil psychique groupal: constructions du groupe. Kaës was aware of Foulkes’ work; he shared the idea not only that a group can be understood as a whole, but also, and especially, that the group environment takes on an unconscious dimension (Kaës, 2015: 78). However, he found that Foulkes had not specified precisely what is involved in the unconscious character of the group and its members (Kaës, 2015: 78). Buildling on the work of Spitz, namely his notion of three basic organizing principles (‘organizers’) in the psychological development of the child: the smiling response, the stranger anxiety and the origins of semantic communication (Spitz, 1965), Kaës argues that the group—and, more broadly, the group bond or link—can be organized around an effective fiction, that of the group psychic apparatus. The function of this apparatus ‘is to transform and bind the psychic formations of the members of the group by enlisting the aid of a structural unconscious organizer (organisateur), such as psychic groups’ (Kaës, 1976: 204). Kaës shares Foulkes’ idea that the members of a group are in a state of interaction. He adds that these interactions are punctuated by a double movement that is the source of the group formation: the struggles against both the death drive and the anxiety of not having a place (Kaës, 1979: 205). These interactions are not, however, organized into the form of a network; instead, they can be divided into three interdependent dimensions, which themselves constitute the group psychic space. These three dimensions are: 1) the intra-psychic, the inner space of the singular individual or subject; 2) the inter-psychic, which concerns the psychic space between subjects in the different configurations of their bond (couple, group, family, and the subjectivity-effects of each of these); 3) the trans-psychic, which is the psychic reality that is transmitted through the subjects (Kaës, 2015: 67). As can be seen in Figure 1, Kaës does not view the psychic space of the individual in the group as a crossing point in a web of complex communication; instead, it is a place that is encased within and interdependent on other spaces.
The three overlapping dimensions of the space of psychic reality (Kaës, 2015: 67).
Both Kaës and Foulkes hold that the individual’s positioning within and participation in a group, as well as all of his/her attitudes and stances, express a specific group dynamic (Kaës, 2013). Kaës, however, explains that this dynamic is brought about by the activation of specific organizers within the group: the unconscious psychic organizers. This involves complex formations that ‘sustain and express the comprehensive development of group bonds’ (Kaës, 1993: 178). He then distinguishes between the organizers of the intra-psychic dimensions of group space and those of the inter- and trans-psychic dimensions of this space, such as the body image, elements of primal phantasy, family complexes, and imagos (Kaës, 1979: 26).
The matrix and GPA models: two complementary vertex
The matrix and GPA models are genuine metapsychological proposals that provide different perspectives on how the group environment is formed and structured. This is why we conceive of the two models as vertex, as Bion (1965) defined this term. The matrix and GPA models have distinct ways of understanding the group: the former considers it to be a system of communication and the latter holds that it is an unconscious psychic formation. Yet this does not tell us how the matrix and the GPA models can be related to one another.
To take up this question, it is useful to turn to Devereux’s (1967, 1978) ‘complementarist’ approach. Devereux’s understanding of the complementarity relationship, or double discourse theory, can shed new light on two approaches or scholarly disciplines that have been separated only by established disciplinary boundaries. Complementarism affirms that the human being, as an object of study, is a unity; the different autonomous and irreducible fields or explanations of this object are complementary if they are neither coextensive nor in competition with one another. The matrix and GPA models are two descriptions of the group that were written during different periods and are based on different clinical material. They are not, however, rival theories that compete with each other; it is much more productive to view them as complementary. Foulkes worked out his idea of the group as matrix in relation to his work with groups in Birmingham during the Second World War, in the second of what were known as the ‘Northfield Experiments’; Kaës explored his intuition of the group as psychic apparatus in the 1970s, while working with training and psychodrama groups at the Cercle d’études françaises pour la formation et la recherche active en psychologie (Circle for French Studies on Active Research and Training in Psychology). The members of the groups that Foulkes observed and conducted were hospital patients, while those in the groups that Kaës led were usually healthcare professionals, who were involved in continuing education in group therapy.
The matrix model accounts for the various ways that the members of a group link up and, especially, communicate with one another. The concept of the network, which is central to this model, can be grasped in terms of four dimensions: 1) the network is the structural web of a group, providing it with a topography; 2) it is the locus where individual and group psychic processes emerge; 3) it is the field in which the members of a group interact; and 4) it constitutes all of the channels through which they communicate. Pines (1991) and Hinshelwood (2009) have highlighted how little this model has changed since it was introduced. The GPA model, on the other hand, provides a way of constructing a new metapsychology not only of groups but also of the psychic function of the subject (Kaës, 1976).
Conceptualizing networks in the matrix as ‘internal groups’ and ‘unconscious alliances’
We find that the metapsychological features of the GPA model—in particular the ‘internal group’ (Kaës, 1976) and ‘unconscious alliances’ (Kaës, 1994, 2009)—provide a way to conceptualize the emergence and dynamic structuring of communications networks in the matrix model.
According to Foulkes (1990: 212), ‘The group, the community, is the ultimate primary unit of consideration, and the so-called inner processes in the individual are internalizations of the forces operating in the group to which he belongs’. In 1976, Kaës provided a complementary proposition: the individual is structured like a group. Referring to Freud (1921), Anzieu (1966), and Pontalis (1963), Kaës (1976: 192) posited that the individual psychic apparatus not only has group properties, but is also composed of group substructures, such as ‘internal groups’. Kaës (1993: 130) defined the internal group as ‘intra-psychic formations and processes, from the point of view in which the relations among the elements that constitute them are organized by a group structure’. This concept led to new ways of thinking about the psychic life of the individual within the group setting. There are different types of internal groups, a term that includes the body image, primal phantasies and Oedipal and fraternal complexes (Kaës, 1993: 135). Finally, in intra-psychic environments, internal groups allow the organization of the GPA in the group situation: they ‘provide it with its basic structure through projection, projective identification, adhesive projection or incorporation, and displacement, condensation and diffraction’ (Kaës, 1976: 235). We would like to propose that the individual in the group setting—a setting that is caught up in the matrix’s network of communication—is fundamentally structured like a group and is composed of internal groups. From this perspective, the communication networks of the matrix could be thought of as the externalized expression of the deep structure of the individual; the latter is already a group structure that is made up of internal groups, which can then be disclosed by the group situation.
We also find that Kaës’ notion of unconscious alliances provides us with a way to describe what underpins the organizational logic of a group’s communication networks. In other words, the logic of a group’s organization can be partially understood, at the unconscious level, in terms of the alliances that have been set up between the individual members of the group. Kaës (2009: 35) drew on Bion (1961) in defining unconscious alliances as unconscious formations that are shared by all the group members. Unconscious alliances, which are entered into by both the subject and the entire group to which s/he belongs, emerge in two distinct spaces: the intra-psychic and the inter-psychic (Kaës, 2009: 78). There are two types of these alliances: those that provide a specific defence and those that carry out the unconscious desires. Thus, Kaës (1994, 2009) distinguishes four groups of unconscious alliances: primary structuring alliances (arising from the mother–child bond), secondary structuring alliances (formed by contracts and pacts arising from fundamental taboos), meta-defensive alliances (the individual mechanisms of the group members are taken over by the defences constructed by the group itself) and offensive alliances (these allow a coalition to be formed in preparation for an attack).
Starting in 1964, Foulkes also drew attention to different modes of interaction between the members of a group and the therapist. For example, on observing the level of conflict rising during a group meeting, Foulkes (1984: 119) identified several group phenomena, such as displacement, which refers to a tendency that has been repressed in one individual and appears in the role played by others; isolation, which involves attributing to one group member characteristics that have been phobicly avoided by the others; and splitting, which can manifest itself in the creation of sub-groups (Foulkes, 1964: 119). Yet Foulkes does not provide a model that explains how such phenomena emerge and, especially, how they are organized within the network. Because unconscious alliances allow specific defences to be maintained and unconscious desires to be realized—while also organizing the linking that occurs in inter-psychic space—they can serve as an effective means of explaining the underpinnings of the phenomena described by Foulkes. The phenomena he observed can thus be thought of as expressions of the unconscious psychic alliances that have been established by the members of a group.
In the following clinical illustration, we shall use these complementarist ideas to try to understand the dynamics of a group of professionals participating in group supervision.
From the expression of a psychic alliance to the restructuring of communication networks: a clinical example
One of the present authors was asked to work with the employees of a youth centre 1 in south-western France, and to provide them with a space to talk about and reflect upon both their working methods and the young people they served. After the preliminary interviews with team members and administrators, they were given the opportunity to take part in a monthly supervision group that would make use of role-playing. This group was open for all to attend. During these sessions, the group’s supervisor invited members to use free association and role-playing in order to facilitate the sometimes difficult task of putting into words their experiences of the concrete situations they encountered each day, whether these involved their colleagues or the young people with whom they worked.
Each supervision session was set up as follows: participants were seated in a semi-circle and asked to use free association when speaking about the problematic professional situations they had encountered. Then, in a second phase, the supervisor asked the group to pick one of the situations that had just been mentioned and use it as the basis of a role-play. Next, in phase three, the members who had been chosen to participate in this role-play would act out the scene before the semi-circle of seated members. At the end of the scene, the actors were asked to take their seats and the session closed with a sharing of the thoughts and feelings that had been brought out during the role-play; the actors and the other members of the group would process the difficulties that had been experienced during both the role-play and the initial situation. The group interaction was sustained by the rules that the participants had agreed to follow: confidentiality, regular attendance of sessions, punctuality, and respecting other group members and what they said.
Phase one: from group abandonment to overstimulated communication networks
The first session of the supervisory group was to be attended by Juliette, Sarah, Olivia, Maria, Marion, Fanny, Aurélie, Pauline, Muriel, Solenne and Paul. When the members arrived and took their seats, however, it was apparent that one person was missing: Muriel, the manager who coordinated the youth centre. None of the others knew why she was absent; her place would remain empty throughout the first session. However, the supervisor and the administrators had agreed that everyone would be expected to be present, regardless of his/her position in the hierarchy of the organization.
Paul was quick to speak: So, I’ve been having a problem recently with one of the kids, Younes. We were walking out of a discussion group with some other kids and he suddenly got very upset. I thought he’d completely lost it—he was threatening to kill himself and the rest of us, as well. It was borderline delusional. I had to send the others away and keep Younes by himself until he calmed down. The supervisor asked if Paul had any ideas about why this had happened, and he replied, ‘He’d been told that he was no longer eligible for benefits’. Paul added, He came back to the centre the next day, when I was on my break. I didn’t want to see him in my office, as a matter of principle. But it was complicated. He wanted to call his father in Martinique to ask him to send him some money, but I was afraid he’d freak out again, and meanwhile the others had already gone into the office, so that they wouldn’t be with us in the corridor. While Paul was speaking, the other group members constantly spoke over him, creating a disorganized din, in which members interrupted each other, asking for more information and details about the youth and the situation.
The supervisor had to ask them several times to let Paul finish explaining the situation, reminding them that they had agreed to respect one another’s speech. The supervisor noticed that the chain of group association was especially dense and was laden with multiple comments, which saturated the mind in a continuous stream of words and provoked a slight, countertransferential sense of nausea. Paul was finally able to conclude: ‘Actually, I wonder what I could have done to keep him from flipping out like that and to protect the others’. ‘Keep someone from decompensating’, ‘protect the others’, ‘protect oneself’: Paul’s words resonated in the supervisor’s thoughts, becoming a request on the part of the group to be protected from something. The supervisor realized that reminding the group to respect everyone’s speech could be seen as the necessary expression of a framework, one that could protect them from the anxiety that had seemed to permeate their initial exchanges.
In the first moments of the life of this group, we can already see that the group’s chain of associations had become saturated, and this situation was maintained by the overstimulation of the communication networks. The members interacted incessantly during the first period of their exchanges in intersubjective space: everything Paul said was questioned, associated to, and commented on by the other participants in a cacophonous movement. It was as if whenever anyone made any statement, the others could not prevent themselves from generating an uncontrollable chain reaction within the networks of the space of linking. How can this overstimulation of communication networks be explained?
As everyone knows, the moment when a group is formed is a delicate phase of the group dynamic. Between the increase in the porous quality of psychic envelopes (Vollon et al., 2014) and the emergence of intense anxiety, the moment of group formation can generate significant movements of defensive manic response and even conflictual tendencies (Foulkes, 1964). Here, this formation was marked by an important event: the unexplained and unexcused absence of Muriel, who was expected to be there like the rest of her team. Her absence may have brought into play certain paranoid schizoid anxieties (Klein, 1946) connected with the loss of a part of the group during the very first moments of its meeting. Paul had chosen to present a situation that involved a young man who had been abandoned by the social welfare system and left, estranged from his family, in a vulnerable and precarious position. This choice can be understood as the group’s representation of its members’ feelings of anxiety about the disappearance of part of this group. In other words, they felt that part of the group’s body had been lost. Viewed thus, the relatively chaotic interaction among participants and the consequent overstimulation of its communication networks—none of which was contained by the rule of respecting what others say—expressed a double movement. First, the group members set up a meta-defensive unconscious alliance among themselves by saturating the group chain of association and not respecting the rules of the supervision; this enabled them to distance themselves from their supervisor, who was prevented from intervening. This impediment is perceptible in the supervisor’s countertransferential feelings: the nausea, which represents the overload of words, served as a screen that affected the way in which the supervisor listened. By forming this meta-defensive unconscious psychic alliance, the members of the group kept the supervisor out of things; they did so in order not only to protect themselves against having to put the feeling of archaic loss to work in the session, but also to give themselves a feeling of unity that would enable them to deal with Muriel’s absence.
The overstimulation of the communication networks can also be explained by the content of Paul’s comments. In focusing on his own inability to contain the young man’s agitation, Paul allowed a form of linking—affected and shaped by anxiety and by a fear of the expression of destructive drives—to circulate from his intra-psychic space to the inter- and trans-psychic space of the group. In other words, it was as if Paul’s own internal psychic ‘groupality’, which was characterized by his agitation, disquiet, and vulnerability before the fragile and unstable young man, had permitted the emergence of a structure of communication networks that were themselves similarly agitated and overstimulated.
Phase two: the restructuring of communication networks
The supervisor decided to remind the participants of the rules, both to contain the overstimulation and to enable the group members to form inter- and trans-psychic links; in this way, they could organize their communication networks. The supervisor thereby provided a two-sided intervention, which touched on both the form and the substance of the group as a whole. The supervisor said: This supervision and its rules—especially the rule about respecting what others say—are a new way for you to be together with each other. It may upset you that Muriel isn’t here, but we can think and talk about things together in a respectful way, even if some of us are absent. The group remained silent. Then Fanny began to speak: OK, right now, I’ve got a kid who’s been radicalized. He was in jail but he’s looking for a job now. Everyone gets upset about him: he’s tall, well-built, intimidating. No one knows his family. He’s all alone in the world. During our first session, he refused to shake my hand because I’m a woman. All he did was ask for things—a lot. This time, the group was attentive while she spoke. Juliette chimed in to speak of the loneliness she felt in her practice. This led the supervisor to think of the absence of the centre’s coordinator; the image of an orphan group formed in the floating attention of the supervisor, who decided to ask the group to role-play the first session between Fanny and the young man she had described. Fanny decided to take the role of the young man, and Marion volunteered to play Fanny. The supervisor suggested that they make up a name and an age, in order to enhance the sense of improvization and ‘make-believe’ in the role-play. Fanny would be ‘Abdel’, aged 18, and Marion decided to call herself ‘Olga’ and be 45 years old.
During the role-play, ‘Abdel’ was indifferent and demanding. He asked for money to buy some clothing. ‘Olga’ started to lose patience: ‘Fine, but you know, I have to work if I want to buy things for myself. They don’t just appear from nowhere like magic’. The role-play ended and the two women returned to their seats. As it was being acted out, the supervisor felt a sense of exasperation and even hostility towards ‘Abdel’. These same feelings were present when the group members discussed the role-play. Everyone agreed that ‘Abdel’ was completely unbearable and that lots of the kids at the centre were ‘like that’; they would make demands that would sometimes oblige the group members to step out of their professional roles and provide assistance ‘for a little bit of anything and everything’. The supervisor thought of a nest filled with baby birds waiting with open beaks to be fed.
In this second part of the formation of this group, the networks of communication were being restructured: the overstimulation had calmed and the communication flowed in a more relaxed manner. The group’s lived experience of abandonment and vulnerability had found a way to be enacted both through Fanny’s description of her sense of being overwhelmed, and through Juliette’s statement that she felt alone in her work. This restructuring was then represented in the role-play: the ‘Abdel’ figure, with his expression of his many needs, can be viewed as a representation of the overstimulation of the communication networks. More broadly, this figure also represented the group itself, since it had been orphaned by the absence of its coordinator and had completely filled the space for speech with its own numerous remarks and demands. The ‘Olga’ figure can be understood as representing the frame and the rules of the group process, which the supervisor had reminded them of in the second part of the session.
By way of conclusion
In continuing with our project of bringing together the matrix model of the group with the models developed by psychoanalysts in France, we have sought to provide a different way of reading the work of Foulkes and that of Kaës, which has received less attention in the literature published in English. Bion’s (1965) concept of the vertex and Devereux’s (1978) complementarist approach have enabled us to formulate several proposals about what may underlie the formation of communication networks and how to understand the forms of interaction identified by Foulkes (1964). This research is still in its preliminary phase. Both Foulkes and Kaës had a rich and complex understanding of the group and its processes; our clinical example emphasizes our need to advance further in our effort to connect the two models. Various questions remain: what are the epistemological limits of the use of the idea of unconscious alliance to understand the underpinnings of the modes of interaction in the matrix? What links and connections can be made among the models of Anzieu’s envelope, Foulkes’ matrix, and Kaës’ GPA? We plan to address these and other questions in future publications.
Footnotes
Acknowledgements
Translated by John Holland
