Imagining the Other: The Influence of Imagined Conversations on the Treatment Process

Preface

The conduct of psychotherapy is a complex process. Therapists constantly question which topics must be revisited, which must be processed in a different way, when new issues arise, and how to answer an infinite number of other such questions. Process research is concerned with how therapists make these decisions within a treatment. This study suggests that one method some therapists use to guide their practice is their imagination.

From Freud (1907) to Coleman (2006) there is a strong current among analysts to view imagination as a sign of ego weakness and a turning away from reality. Yet, the participants in this study attest to the fact that it can be a powerful problem-solving tool to move past difficulties in the psychotherapeutic treatment process. The participants of the study demonstrate that imagination can aid in the executive functions of analytical thinking and the application of theory to help guide treatment. Likewise, they express the degree to which imagination does not arise in a single mind, divorced from reality. Rather, their use of imagining clients is so entwined with the memory of intersubjective interactions that they are able to use the imagined conversations to deepen their understanding of their clients.

The experiences articulated by the participants in this research are not unique to psychotherapists. Indeed, neuroscientific research on imagination supports the claims that imagination involves memory to a very high degree and follows the rules of analytical reasoning. This is undoubtedly why the neuroscientists report that imagination aids in problem-solving. The data from this study suggest that imagination stimulates problem solving ability due to activity in mirror neurons, and these in turn, help therapists develop a greater understanding of clients. That is, the therapists can use imagination to surface their implicit relational knowledge of the client to understand the client’s intentions, motivations, feelings, and meaning, and hence, adjust treatment to fit the client’s needs.

Introduction

Process researchers have long investigated what happens during therapy (Schroder, Wiseman and Orlinsky, 2009). A few studies are now showing that experiences both the therapist and patient have between sessions regarding treatment may also be of great importance (Hartmann, Orlinsky, & Weber, et al., 2010). Researchers have termed these “intersession experiences,” which are defined as

While most of the work in this area focuses on the patient’s intersession experiences, one study in particular looked at the intersession experiences of therapists (Schroder, Wiseman, & Orlinsky, 2009). Among the intersession experiences these researchers measured was that of therapists imaging conversations with clients. More than half of the therapists they surveyed reported imagining a conversation with clients between sessions.

These findings suggest that the experience of having an imagined conversation with a client between sessions may have some importance to the therapeutic process—and perhaps even to the outcome. It would appear that more information about this phenomenon is desirable. To that end, the researcher undertook a qualitative study to elicit psychotherapists’ descriptions and opinions regarding intersession imagined conversations with clients. The general nature of the question suggested that a Constructivist Grounded Theory (CGT) method might yield information about this phenomenon.

In keeping with the spirit of CGT, the definition of “imaginary conversation” was left open on the questionnaire distributed to the participants. Rather than imposing the meaning of the term on the participants, the researcher sought to have a definition of the phenomenon arise from the experiences of the respondents. The researcher then turned to the literature and was surprised with the degree to which the respondents’ experiences matched the understanding of imagination proffered by neuroscientists.

Review of the Literature

The role of literature in CGT is an issue of some concern. One tenet those who have developed and employ CGT agree upon is that the literature should not exert a constrictive influence over the data analysis (Mills, Bonner, & Francis, 2006). For that reason, the literature was consulted after data analysis had begun. Thus, the reported experiences of the therapists guided the literature review, rather than the literature guiding the analysis of the data. There are three bodies of literature that address imagination in some depth: philosophy, psychoanalytic theory, and neuroscience. A full review of all of these sources is beyond the scope of this article. The researcher inspected these bodies of literature after reviewing the bulk of the participants’ responses and found that these reports were quite consistent with findings from neuroscientific research. The findings of this study are also supported by the work of Fonagy, Gergely, Jurist, and Target’s (2002) work on mentalization or reflective functioning. Thus the literature review will focus on these areas.

Fonagy et al. (2002) drew upon a vast array of sources—not least of which is the field of infant research—in formulating their understanding of mentalization. They also refer to “reflective function,” which they define as the “ operationalization of the mental capacities that generate mentaliztion…” (p. 3). Perhaps the most concise definition of mentalization was offered by Allen, Fonagy, and Bateman (2008). They identified the concept as “holding mind in mind” (p. 3). That is, it is the ability to determine that one’s own and another’s experiences are mediated by mental states as well as the ability to conceptualize those states. It includes the ability to “read” another’s mind to render his/her behavior meaningful and predictable (Fonagy et al.). As elaborated in the data analysis in this paper, imaginary conversations (ICs) appear to help therapists mentalize clients.

The neuroscience literature uses the terms “mental simulation” and “counterfactual thinking” to refer to experiences of the type the therapists in this study reported. Mental simulation is “the act of imagination and the generation of alternative realities” (Markman, Klein, & Suhr, 2009, p. vii). Counterfactual thinking is identified in this way:

Many neuroscientists who have studied imagination concluded that it is an effective problem-solving tool. Research by Faude-Koivisto, Wuerz, and Gollwitzer (2009) demonstrated that mental simulation is an important factor in goal attainment. This occurs through the process of switching on an explorative mind-set in the pre-actional phase of problem-solving, which results in the generation and consideration of alternative actions. Their research has shown that using imagination as a problem-solving device “enhanced performance on creative generation tasks and led to more expansive information processing with broader conceptual attention” (p. 73). They concluded that imagination as a problem-solving strategy is best suited to novel or complex goals.

Byrne (2007) demonstrated that counterfactual thoughts utilize the same principles as rational thoughts. In fact, Byrne stated that rational thought, including conditional reasoning and causal relations, is premised upon imagination (see also Byrne and Girotto, 2009). In support of this hypothesis Gerlack, Spreng, Gilmore, and Schacter (2011) used functional Magnetic Resonance Imaging (fMRI) data to demonstrate that mental simulations aimed at problem-solving extensively involve executive regions of the brain. These simulations require that one “formulate a plan to solve the problem, integrating and sustaining relevant information, and maintaining an abstract sequence of steps leading to the problem’s solution” (p. 1821). In other words, imagination in the service of problem-solving can be quite analytical.

Another point arose as important in the data analysis: Imagination involves a great deal of neural overlap with memory. Thus, imagination is more closely related to the perception of the external world than some writers have argued (Schacter and Addis, 2009; Arzy, Collette, Ionta et al, 2009). Szpunar, Watson, and McDermott (2007) conducted a study in which participants were given the tasks of envisioning a personal a future event, remembering a past personal event, and imagining an event that involved someone other than the participant. The participants underwent fMRI during the tasks. The data revealed that while the regions of the brain that were activated differed somewhat during the tasks, there was a significant overlap in neural functioning when a participant remembered a personal past event and imagined a personal future event. The researchers attributed the finding to the theory that imagination is formed by reactivating previously stored images from familiar visual-spatial contexts stored in the episodic memory.

Schacter and Addis (2009) support this theory based on their own research. Their studies led them to formulate the constructive episodic simulation hypothesis in which they suggest that “past and future events draw on similar information stored in episodic memory and rely on similar underlying processes; episodic memory supports the construction of future events by extracting and recombining stored information into a simulation of a novel event” (p. 111). In other words, imagination is based on memory. Because memory is created from the perception of and interaction with the “real world” or “world out there,” imagination is not the private creation of an individual mind. One may extrapolate that in the therapeutic context it is the product of client-therapist interaction.

Methods

The question that propelled this study is: What is the nature of therapists’ experiences of having imagined conversations with clients outside of the therapeutic session? Constructivist Grounded Theory was used to explore this question. Constructivist Grounded Theory is a qualitative research orientation rather than a strictly prescribed set of methods. This theory is comprised of two processes. First, the researcher gathers narrative data. Second, based on this data, the researcher constructs a theory that explains the data. The theory construction may include a literature review, but the literature is secondary to the data itself. The theory thus constructed is judged by the degree to which it offers a coherent explanation of the data, in keeping with available evidence and reasoning.

From a constructivist perspective, following a protocol that was devised independent of the research question cannot enhance the validity of findings (Mills, Bonner, & Francis, 2006). Therefore, there is no template for the collection and analysis of data (Charmaz, 2000): methods are chosen on the basis of the method’s likelihood of helping the researcher both stay faithful to the participants’ experience and build theory (Mills, Bonner and Francis).

This study received exempt status from the institutional review board at the university at which the research took place. Potential participants were informed that their responses were going to be used for research, and all identifying information was removed from the data as it was submitted. The sampling plan was a snowball plan. The inclusion criterion was that the participant must be a practicing psychotherapist. The “master clinician” sampling procedure frequently used in process research was not pursued, as the goal was to gather the experiences of “ordinary” clinicians. The self-identified theoretical orientations of the participants appear in Table 1.

After reviewing the first 12 responses, it was determined that theoretical saturation was reached after participant number seven. Thus, data collection ended after response number 12.

The researcher asked the participants to submit a written abstract, general description of the experience of having imaginary conversations (ICs) with clients between sessions, and/or a specific example of this experience in which the participant engaged. The participants were asked to describe when ICs typically occurred, the aspects of the client and therapist that were included in the phenomenon (e.g. vocal qualities, gestures, facial expressions, etc.), if the participant believed that the IC impacted the therapeutic process (and if so how), and any other information they cared to share about the experience. The data were analyzed using constant comparison, line-by-line, open, and axial coding (Charmaz, 2000). From this process categories and relationships between categories emerged. The fit of theoretical categories to the data was determined by consistency of judgment across observers. Two independent observers read 10% of the data, and had a 74% agreement rate.

An important issue in Constructivist Grounded Theory is the researcher’s bias (Mills, Bonner, and Francis, 2006). As much as one may wish to approach the data as a table rasa, one cannot unlearn what one knows. Over the course of an academic career one pursues an area of interest, conducts his or her research in this area, and consumes the research of others. This forms a lens that the researcher cannot put aside. Recognizing both the biased perspective that the researcher brings to the study and the theories and research that have contributed to the researcher’s world view is imperative.

Having recognized that researchers all bring a perspective to their work, it is incumbent upon the researchers to attempt to disclose that perspective to the best of their ability (recognizing that some bias operates outside of the researcher’s awareness). This researcher is, first and foremost, a psychotherapist. As such she has been interested in the psychotherapeutic process, including questions of how clinical judgment, the therapeutic alliance, and intersubjectivity guide the therapeutic process.

Results

The data analysis in this study uncovered two major perceived benefits of ICs for the participants. One result is that ICs helped therapists uncover countertransferential feelings (defined from a relational-intersubjective perspective; see, for example, Maroda, 2004) relative to the client and/or treatment process (these results are reported in a separate article, Arnd-Caddigan, in press). The second finding, reported herein, was more surprising: The participants felt that ICs helped them understand their clients better. Understanding why/how they believed this constitutes the grounded theory elaborated in the discussion. The building blocks of this theory are the categories that emerged during the data analysis. These include the aspects of interaction that the participants noted relative to ICs: the point during the treatment process when ICs tend to occur most, the contents of the ICs, and the participants’ perceived value of ICs.

Aspects of Client/Interaction in ICs

In response to the question “what aspects of interaction do you imagine?” several respondents described visual, auditory and affective aspects of interaction. One participant described a specific IC:

Another participant gave an elaborate description of an IC with a particular client:

At another point in the narrative the same respondent stated, “Bob’s tone of voice changes to excited or mock frustrated as he bounces forward on the couch. His arm movement is “large,” or far away from this body, rapid with a high degree of muscle tension…”

Still another participant made clear the feeling aspect associated with ICs:

Thus, one sees that facial expression, body movement, and affective tone or general “feel” of interactions tend to be incorporated into the therapists’ imagined conversations with their clients.

When ICs Occur

In response to the question of when in the treatment process ICs tend to occur, there were two answers: 1) when a difficulty occurs and 2) during the middle phase of treatment. The participants in this study overwhelmingly reported that imagined conversations with clients between sessions are a useful tool in helping them problem-solve when they experience a difficulty in the treatment process. For instance, one respondent noted that ICs occur “when I feel stuck with a case.” Another participant stated that ICs occur

The respondents in this study also indicated that ICs occur most frequently in the middle phase of treatment or after a strong therapeutic alliance has been established. One participant in particular elaborated on the importance of this phase of treatment for ICs:

Contents of ICs

ICs are, as the definition of mental simulation implies, hypothetical. Yet, the respondents in the research disclosed the degree to which these experiences include both analytical thought and memory. Analytical thinking is “the abstract separation of a whole into its constituent parts in order to study the parts and their relations” (Free Dictionary.com). One finds the intersection of both analytic and theoretical thinking with imagination in the data provided by the participants of this study. One participant in particular demonstrated this principle in some detail. This participant discussed beginning one IC with identifying relevant themes from the course of therapy.

The therapist has illustrated how he or she pulled out pieces from past interactions to look at their relations to make sense out of the complex information gathered over the course of treatment. These bits, previously learned about the client, informed a question the therapist posed in an IC. Once the therapist imagined the answer to the question he or she reverted to analysis and theory application to arrive at what he or she understood to be a core conflict for the client.

Another participant described the interweaving of analytical thought, theory, and ICs in this way:

When asked what aspects of the client and therapist the participant incorporates into an IC, this same participant’s response was to include “patterns within our therapy… [the client’s] way of dealing with conflict, and ways of relating in general.” In both of these quotes one sees the process of analytical thought: the logical coherent thought process of breaking down the entirety of the client’s presentation into constituent parts to find the logical coherence, alongside the application of theory.

Participant 12 was more explicit about the relationship between theory and ICs:

For this respondent theoretical engagement triggered the IC. The participant then expounded on the interweaving of IC and theoretical analysis:

For the participants in this study, analytical thought, the application of theory, and ICs are inextricably united to generate ideas about what tactics may be helpful to break a therapeutic impasse.

The respondents in this study also indicated the degree to which they employ memory in the process of an IC. Indeed, the data in this study suggest that ICs often flow directly out of memory. Several participants discussed this experience. One documented an actual discussion with a client as the therapist (in the fashion of a process recording) remembered it, then the experience moved almost imperceptibly from memory to an imagined alternate dialogue. During the IC the participant recognized a switch back to a visual memory of the client, noting that the line between imagination and memory was blurry: “I see him [the client] again in my mind’s eye. This is more memory than imagination.” The two phenomena appear here to be merged.

Another participant stated “At first, this reflection process is my recalling the content of the session,” and then slips into an IC. Yet another participant stated it in this way:

Thus it appears as though ICs in the context of psychotherapy are not entirely created by a process that is internal to the therapist. The therapist often begins the process of imagining a conversation with a client by remembering a previous conversation and then glides back and forth between memory and imagination. The imagined part of this process borrows greatly from remembered interactions, to the degree that the two blend into a single experience.

A Value of ICs: Solving Problems by Enhancing the Understanding the Client

Several participants stated that one value of ICs is helping the therapist develop a deeper understanding of the client’s perspective. One participant suggested the value of ICs is in that… “it gives me a chance to really look at what’s ‘going on’ in the person’s mind.” Another participant noted: “They [ICs] also help me understand the fears that a client might be experiencing. They help me sharpen my sense of empathy and understanding.” Yet another participant discussed this benefit of ICs at some length:

In this lengthy excerpt one may apprehend the degree to which the therapist’s awareness of the client’s subjectivity is much more detailed, articulated, and applied to specific content than the account of knowing the client based on phase of treatment alone. In discussing the value of the IC, one sees the degree to which knowledge about the client that is accumulated during the course of treatment is later brought into clear awareness to be analyzed and used for treatment. The participants believe that the IC gives the therapist insight into how the client will experience possible intervention strategies. This predictive capacity then informs treatment decisions.

Of special note in this context is the data supplied by the participant who narrated an IC as it was occurring. There was an addendum to the narrative in which the participant discussed bringing notions that arose during an IC into the next actual session with the client. In the IC the therapist “asked the client:” “How do you feel about being in charge of the session?” to which the client responded, “I feel like you’re not there; like I’m alone in this.” In the actual session that followed, the therapist asked the client the same question. The therapist documented this session: “I asked how it feels to be in charge of our sessions. Bob said he doesn’t want it to feel like I’m not there (I guess I was pretty close in my imagined conversation).” Indeed, the therapist, in this instance, appears to have had access to the client’s meaning system during the course of the IC and had accurately predicted the client’s response to a specific topic. That is, the IC allowed the therapist to bring enhanced awareness to the session and to apply a more generalized relational knowing to a specific treatment problem. This in turn allowed the therapist to use this knowledge to raise an issue germane to the client’s growth.

Discussion

The idea that an act of imagination could potentially be helpful when the therapist is “stuck” (or needs to generate further understanding of a case) is supported by the neuroscientific perspective. What the neuroscientific research does not directly address is why the therapists in this study felt that imagining a conversation with their clients would give them greater understanding into the clients’ subjective experiences. This finding may be counter-intuitive to those who see imagination from the vantage point of a one-person psychology: If it is an experience that is generated from the internal experience of the subject, it cannot yield accurate information about an object: from this perspective it could be no more than a projection. Yet, the neuroscientific research indicates that perhaps imagination is not a one-person phenomenon. If imagination is constructed from memory of the external environment, then there must be some role for the object in the formation of the imagined event. Here conjecture based on both the neuroscientific research and in vivo clinical process research conducted by the The Boston Change Process Study Group (BCPSG)may suggest an answer. The link involves the role of implicit relational knowing.

The theory put forward in this study is that mirror neuron activity is central to ICs. Mirror neurons are structures in the brain that respond nearly identically when an individual performs a purposeful activity or experiences emotion and when he/she watches (or hears) another perform a purposeful activity or display an emotion (Rizzolatte & Sinigaglia, 2009; Braten, 2007). Mirror neurons are motor neurons (Rizzolatte & Sinigaglia, 2009). When an individual watches or hears another perform an act or express an emotion, the observer’s mirror neurons prime the observer to perform the same motor activities. This priming action may lead to the execution of the motor activities, or remain only potentially evoked with the overt action inhibited (Rizzolatte & Sinigaglia, 2008).

Several authors (Rizzolatti & Sinigaglia, 2009; Decety & Stevens, 2009; Gallese, 2009) have stressed that based on the functioning of mirror neurons, understanding another person’s motivations, intentions, affect, and meaning is no longer seen as a vicarious experience. Rather, it is a first-person, non-mediated experience:

The point is worth stressing: The activity of mirror neurons gives the observer an understanding of the target’s subjective experience because the observer shares the experience at the level of overt or inhibited motor activity. By accessing one’s own feeling-state, it is possible to gain awareness into what another is experiencing. This concept is at the heart of intersubjectivity (Beebe, Knoblauch, Rustin, & Sorter, 2005).

Several researchers (Trevarthen, 2009; Boston Change Process Study Group, 2010) have suggested that the knowledge of another’s motivations, intentions, affect, and meaning based on mirror neuron activity is experienced as “a feel” for what it is generally like to be with a specific other person. The feeling is vague, as the knowledge of the other resides on an implicit level and, therefore, is difficult to represent symbolically (BCPSG; Pally, 2005). The BCPSG has termed the feeling “implicit relational knowing” to highlight the degree both to which this information is tacit, rather than existing on a more conscious level, as well as the degree to which one can develop an intimate knowledge of being with another.

Implicit relational knowing may be derived from the mirror neuron response to auditory and visual cues (Pally, 2005). Among these cues are vocal qualities, including pitch, volume, and rate, as well as facial expressions. These latter may be readily notable, or so discrete that they are observable only in frame-by-frame video analysis. In either case they have been shown to impact the subjective experience of conversational partners (Stern, 1985; Bråten and Trevarthen, 2007). Gross body movements are also involved (Beebe et al., 2005). The combined effect of these attributes is the prosody, or rhythm of communication, vitality affect (degree of intensity along with the affective tone), and the shape of the event—crescendo, decrescendo, etc. (Stern, 1985; Stern, Hofer, Haft, & Dore, 1985).

When considering that much of imagination is based on memory, it is possible to conclude that during an IC, one’s own (as well as another’s) non-verbal communication is as much memory as creation. The work of Kosslyn and Moulton (2009) supports this speculation. Their research has demonstrated that implicit memory affects mental imagery. Those aspects of implicit relational knowing almost certainly play the same role as other aspects of memory on the imagined interactions one has with another. In other words, motor memory prompted by mirror neuron activity may help create the motor aspects of an IC.

The question then arises as to how a therapist can use that motor memory during the course of an IC to help find alternative strategies to resolve a difficulty in the therapeutic process. The theory presented is that by re-experiencing remembered motor activity, the therapist’s implicit knowledge of the client may be brought to an explicit level and incorporated into the process of analysis.

This theory is supported by the work of Decety and Grèzes (2006). As stated above, mirror neurons are structures in the brain that respond nearly identically when an individual performs a purposeful activity or experiences emotion and when he/she watches (or hears) another perform a purposeful activity or display an emotion (Rizzolatte & Sinigaglia, 2009; Braten, 2007). The person imagining has similar mirror neuron activity as if actually observing or hearing the target. In other words, calling up the motor memory of being with another person calls forth the first-person understanding one has of another’s experience. The point to be stressed here is that therapists will feel after an IC that they have greater awareness of their client’s subjective experience. The therapists use the ICs to surface previously implicit understanding, stored in motor memory, to apply to the treatment process.

This theory is built upon the premise that imagination is not created from the individual mind of the person imagining: the therapists’ imagined conversations with their clients depend to a large degree on the memory of actual intersubjective interactions between therapist and client. This includes motor memory, which, based on the functioning of mirror neurons, allows the therapist to have implicit knowledge of their clients’ experiences. The ICs function to bring into awareness previously implicit understanding of a client’s subjective experience. This awareness, combined with the use of analytical thinking and application of theory during the IC, can inform treatment decisions to maximize the fit of the treatment to a specific client.

Limitations and Future Research

This study has several limitations and raises many questions for future research. The small sample size was a concern, but it turned out not to be as significant as presumed: Theoretical saturation was achieved within the sample. Nonetheless, a larger sample will undoubtedly yield greater understanding of the phenomenon.

One clear limitation is that (with the possible exception of participant number five) the narrative accounts of ICs were secondary revisions. The narrative represents a remembered experience. It is possible this memory was influenced by a number of factors, including aspects of the therapy that followed the imagined conversation, and does not represent a “pure” account of the experience. Another limitation was indicated in the methods section: constructivist grounded theory is but one possible way to explain the data. It is hoped that competing explanations will be put forward for consideration by members of the therapeutic and research communities.

There are yet other studies that are suggested by this research. What are the differences between therapists who engage in ICs and those who do not? What are the differences between treatment processes for those clients about whom a therapist has had an IC and those for whom the therapist has not had this experience? Does treatment change in a meaningful way once the therapist has had an IC? A close inspection of when, exactly, ICs occur within the treatment process and the client’s perspective on the process before and after ICs would provide potentially useful information. Analysis of an actual ongoing course of therapy would be very helpful in expanding the understanding of the role of ICs in the treatment process.

It may also be productive to examine is the relationship between mentalization and imagined conversations. As mentalization is a developmental achievement, do those therapists who have a highly a developed capacity to mentalize engage more often or more fruitfully in imagined conversations? Conversely, do therapist’s imagined conversations contribute to the development of the client’s reflective functioning? As the capacity to mentalize is based on scaffolding provided by caregivers’ own mentalization of the developing infant (Fonagy et al., 2002), one may speculate that a therapist’s ability to understand a client better based on ICs and use this understanding clinically may propel the reflective functioning of the client. The relationship between the two concepts of mentalization and imagination can be investigated from a multitude of perspectives.

Conclusion

While some research focuses on treatment outcomes, other research is centered on the process: how psychotherapists make decisions and guide the treatment process. Recently this definition has been expanded to include intersession experiences of both the client and therapist. One such experience therapists have is imagined conversations with clients. It turns out that for some therapists ICs may be an important element in navigating difficulties in therapy. The finding of this study is that some therapists use imagined conversations with clients between sessions to break out of impasses. They perceive the effective aspect of the experience to be in deepening their clients’ perspective.

Many clinicians may find the results of this study to be interesting: intersession imagined conversations help some therapists move the treatment process forward. The findings, however, do not explain why this may be so. Thus the researcher constructed a grounded theory to provide a possible explanation. This theory it may be possible that the increased insight into the client’s subjective experience can be attributed to the motor memory or implicit relational knowing that arises as a result of the activity of the mirror neuron system. Given that the mirror neuron system is believed to undergird an ability to understand another in a first-person unmediated manner, putting motor memory into action may well help therapists surface an implicit understanding of their clients. This in turn will help them make treatment decisions that are, in the words of one of the respondents, “safe and beneficial for the client.”