1-5 Vampirism and borderline phenomena: What Carmilla has to tell us about countertransference with borderline structured patients

1-5 Vampirism and borderline phenomena: What Carmilla has to tell us about countertransference with borderline structured patients

No Comments

Jess Dunn


Often, the topic of what the analyst thinks about and feels toward the patient, whether discussed in the context of clinical practice or on the level of theory, cannot be broached without a significant amount of discomfort. This phenomenon called countertransference presents problems in both the theory and praxis of psychoanalysis (and psychotherapy in general). Theorists and practitioners alike are confused with even the most basic questions surrounding the phenomenon: What is countertransference? Is it merely left over bits of unanalyzed unconscious on the part of the analyst that is somehow activated in the analytic situation by particular patients? Can it be understood as some sort of direct communication from one unconscious to another? What place does countertransference occupy in the analysis? Is there a place for it in the analysis? For some theorists and practitioners, countertransference has just as much place in the analysis as transference and can be used to understand and interpret what comes out in the course of the treatment and even, to an extent, guide its course. Other theorists and clinicians, however, remain skeptical about the usefulness of countertransference in the analytic situation. For them, the use of countertransference by the analyst is a mistake, it is a move that distracts and detracts from the true work of analysis, which lies in the patient’s transference and requires analytic neutrality (Campbell, 1982).

Detraction from the work of analysis is not the only or even most problematic potential consequence of the misguided use of countertransference. It is the work that may replace it, the relationship that may develop in place of the analytic situation that presents the greater threat. When the analysis becomes just as much (or even more so) about the analyst’s feelings and thoughts, the boundary between patient and analyst blurs and may even disappear altogether. The feelings and thoughts of the analyst get conflated with those of the patient and, in extreme situations, may be seen as one and the same. The relationship between analyst and patient becomes confused with the type of relationship that one has with friends and lovers, on the part of the analyst and the patient. Ultimately, the fundamental structure of the analysis becomes forfeit, and we are compelled to ask for whom is analysis being conducted. The risk of these sorts of problems arising when countertransference is used as a means of case formulation, interpretation, and the structuring the course of analysis is of no greater importance than when working with borderline structured patients. For patients structured in this way, who tend to vacillate between extremes in both affect and thought, for whom the boundaries between self and other are already so permeable, countertransference that is poorly managed and/or ill-used can be destructive to the patient and the analysis (McHenry, 1994). To elucidate more clearly the particular problems of countertransference with borderline structured patients, I have chosen a rather unorthodox case study. It is the myth of Carmilla: a peculiar and beautiful girl who, through a strange turn of events, was given over to the care of Laura, the lonely and isolated daughter of a petty nobleman in the Austrian country-side (Le Fanu, 1871).

The Case of Carmilla

Carmilla came to be in the care of Laura after she witnessed Carmilla being pulled unconscious from a runaway carriage crash. Her mother, an elderly woman of noble birth was most insistent that Carmilla was in no condition to travel but that she, herself, was compelled to continue without her daughter. Carmilla regained consciousness slowly after her mother had departed and Laura’s father, who had come to the aide of the carriage, transported her back to their secluded mansion. Almost immediately, she and Carmilla became intensely close. They even exchange a childhood dream where each saw the other as young women, as they are now. However, the moments of tenderness, which at times became so intense that Laura becomes faint and even frightened, were punctuated with Carmilla’s sudden enraged or morbid outbursts (Le Fanu, 1871).

The days and weeks passed and Laura began noticing changes in her own mood and level of energy decline. She, like Carmilla, began sleeping late and experiencing sudden shifts in temperament that seem to come from nowhere. Laura started to look and feel tired, wane, and faint and was plagued by visions of a hellish cat-beast that crept into her room at night and bit her chest and then turns into a woman who leaves through and open window. Laura was at once terrified of and deeply attached to her charge and, in turn, Carmilla jealously guarded her relationship with Laura by lashing out at all with whom she shared even the slightest bit of affection or admiration (Le Fanu, 1871).

As Laura became more and more ill, Carmilla more and more possessive. Unable to find an explanation for these symptoms or for Carmilla’s odd behaviors, Laura and her father are compelled to look into Carmilla’s past. These investigations lead to a man whose daughter died of similar symptoms after meeting the Countess Mircalla (an anagram of Carmilla) of Karnstein, who it was later discovered had been draining her of blood and spirit. This name leads them to the abandoned village of Karnstein, where they find the empty grave of Countess Marcilla. At this point, once she had discovered that Laura had been taken from the mansion, Carmilla came in search of Laura and found her at the grave. Upon her arrival Carmilla was killed, it having been determined that only her death could free Laura from this terrible affliction (Le Fanu, 1871).

Borderline Structure

Carmilla first came to them in what appeared to be a fragile state, following a carriage accident that left her faint and confused. Once this initial delicacy had passed and Carmilla was well enough to be introduced to her new caretaker and companion, Laura, her spirits rose immediately. The relationship between the two young women became close almost at once and grew in intensity each day so that, within a few weeks of having met, the two have developed an intimacy that exceeds that which is found in relationships that span years. However, for as charming and beautiful as Carmilla was, she was possessed of certain odd habits and tendencies that even her adoring companion could not ignore. Most noticeably was the mercurial nature of her mood where one moment she was happy and playful and the next she could fly into a fury for which there seemed to rhyme or reason. These strange and sudden changes were also present in her relationship with Laura, where, after a period of cold avoidance or hostility, Carmilla would run to her and shower her with physical and verbal affections so effusive that Laura quite was taken aback. As the story unfolds, it becomes clear that Carmilla wishes for there to be no boundary between herself and Laura. She wishes for a complete fusion between them. She is a vampire, who, in her current state, can only live through attaching herself to the life-force of another (Le Fanu, 1871).1

In the descriptions of Carmilla’s strange behavior it is not difficult to find the same sort of patterns described in individuals with borderline structure or borderline organization as it referred to by McWilliams (2011). The suddenness with which her moods and attitudes towards others can change from one extreme to another parallels the instability that is often central to a diagnosis of borderline personality disorder. To describe these vacillations as instability, however, does not seem to get at the underlying dynamics. In many ways it makes sense to understand this so-called instability to tolerate or even understand ambivalence as a manifestation of those with borderline in its organization. Given that all relationships and experiences fall short, to some extent, of our wishes and expectations, this presents a significant problem for those with borderline structure as they move through the world. Thus, the intolerance of ambivalence in a complex and often disappointing environment necessitates that the individual place everything and everyone into the category of good or bad, a phenomenon that McWilliams (2011) refers to this defense mechanism as splitting, which accounts for the intensity of the emotions and reactions expressed by these individuals. The mercurial nature of these emotions and reactions comes from the tendency for borderline structured individuals to change the way they position others in relation to the good/bad divide with very little provocation. Thus, one moment a person may find him/herself the object of affection and admiration and the next the object of seething hostility and resentment, vacillating between ideas of intense idealization or devaluation.

Similarly, the quickness with which Carmilla became so intensely attached to her new companion is a pattern of relating to others that is often found in and indicative of borderline structure. What is often mistaken initially for warmness and an instant connection between individuals is more than merely liking or getting on well with another person. It is an intense reaction to another which can border on obsession. It is the consequence of an almost complete lack of a sense of boundaries between herself and the other, in which the borderline structured person not only experiences intense affect toward and thoughts about the other, but elicits or attempts to elicit the same intensity from the other in a manner that matches or is complimentary to his/her own. In this gothic tale, Carmilla is all too quick to offer her heart to her new companion and all the hatred and love enclosed therein. Like the borderline structured patient whose initial reaction to the analyst is one of admiration and warm affinity, what underlies Carmilla’s openness and devotion is a desire to at once absorb and be absorbed by the other completely, “to live in your warm life [while] you shall die—die, sweetly die—into mine” (Le Fanu, 1871 p. 22), a desire that may be fulfilled in the analytic situation where countertransference is mismanaged and ill-used.

Countertransference and the Course of Analysis

As was the case with Carmilla, borderline structured patients often come to us in a state of crisis. They come in with tales of a harrowing journey that is brimming with disturbing symptoms, terrible experiences as well as mistreatment and abuse at the hands of others. This is not to suggest that this is never the case for these patients (in fact, it more often than not is) and cannot reflect the events of their lives. However, these accounts also reflect the “bad” side of the good/bad split experienced by the patient, a citizenship that is sometimes ill-deserved. More often than not, if the analyst is at all skilled in establishing any sort of rapport, he or she has the good fortune to be placed on the good side of the divide. In the wake of the patient’s heart-wrenching account coupled with the idealization of the analyst, it is not surprising that an analyst may feel a strong connection to and affinity for this patient in the initial sessions. Similarly, the analyst is also likely to feel that the analysis is going quite well—after all, the patient seems to respond well to them and she is open and communicative (Campbell, 1982). This was certainly Laura’s experience of her new charge and companion, who claimed to be so charmed by her beauty and kindness that she was no longer afraid, confessed a sense that she “already a right to your [Laura’s] intimacy” (Le Fanu, 1871 p. 19). Like the analyst who feels confident in the alliance created with the patient, Laura was hopeful and excited that her relationship with Carmilla was going so well despite the unfortunate circumstances that brought them together. However, in both cases there is misrecognition. What is taken for therapeutic alliance is actually a reflection of the dynamics at work in the borderline structure.

If this misrecognition persists throughout the course of analysis, the analyst will continue to feel that all is well within the analytic situation and even, perhaps, that this is due to their skill as an analyst. The patient will seem quite pleased with the course of treatment and is all too ready to attribute this to the acumen of the analyst. However, given the patterns often observed in borderline structured patients, this idyllic state of things can only persist so long. Sooner or later, the alliance will shift, often without any clear provocation (perhaps the analyst asks a question that the patient found distasteful or announces that she is taking a vacation that will necessitate that the patient miss a session). The analyst who could do no wrong suddenly becomes the analyst who knows nothing, who cannot help anyone, who offers the patient nothing but more pain and distress. The idealization that first marked the view held by the patient about the analyst has been replaced with intense devaluation. And, just as intensely as the previous reaction supported the analyst’s narcissism, the sudden reversal of fortunes wounds it. A puzzling interaction between Carmilla and her caretaker parallels just this sort of pattern. The two are walking along a scenic path together, Carmilla’s hand entwined ardently with Laura’s. They come upon a peasant’s funeral procession singing a mournful hymn. Carmilla is displeased at the sight and sound of the religious spectacle (for reasons still unknown to her companion) and tries to move past quickly, but Laura, out of reverence slows and joins the hymn. Carmilla immediately flies into a rage and proclaims, “What a fuss! Why you must die—everyone must die; and all are happier when they do. Come home!” (Le Fanu, 1871 p. 24). Just as quickly is she entranced by Carmilla’s doting adoration, Laura is frightened, bewildered, and crushed by an unprovoked and vehement outburst.

In the midst of these intense vacillations between idealization and devaluation and the patient’s accompanying emotional highs and lows, it would be unrealistic to expect that the analyst/therapist will not have their own, personal response. However, it is the way in which these personal thoughts and affects that arise in the course of analysis and are taken up by the analyst that will decide whether or not the countertransference detracts from and derails the analysis. When the analyst can recognize these thoughts and affects as a personal reaction to the way in which the borderline structured patient positions the other she can continue the work of analysis via transference. However, if the analyst takes up the countertransference as a direct communication about the relationship, interpretations, or interventions within the analytic situation, then the course of analysis will be altered, almost always for the worse.
In order to make the point more clearly, imagine an analysis with a borderline structured patient wherein the success or failure of the analyst’s interpretations and interventions is determined primarily by the way in which the patient makes the analyst feel. In other words, when the analyst is pleased by the patient’s expressions of happiness with and admiration for the analyst, then the analyst deems the intervention/interpretation successful and the course of analysis to be right. Complimentarily, when the analyst is displeased by the patient’s expression of anger and disapproval toward the analyst, then the analyst deems the intervention/ interpretation to be a failure and alters the course of analysis accordingly. Thus, the good or bad feelings of the analyst, as influenced by the good and bad reactions of the patient come to replace an understanding of the formulation of the case and analytic praxis as the determinant of the course of treatment (McHenry, 1994). Again, we see this precise pattern emerge in the relationship between Laura, who, at the beginning of the tale, shows a healthy curiosity concerning the past of her mysterious charge as well as the odd habits (sleeping only behind locked doors, sudden outbursts of anger, and a vehement hatred of any religious expression) and somatic symptoms (a languid appearance and physical weakness, punctuated by strange trembling fits) she often exhibits, phenomena of equal interest to an analyst. As time passes and she becomes quite attached to the devotion and adoration of her companion and equally fearful of Carmilla’s wrath and dismissal, however, she no longer questions or even outwardly takes note of these idiosyncrasies. Her good, analytic curiosity is replaced by the mere desire to manage her own discomfort by avoiding the anger and hatred of her companion (Le Fanu, 1871).

Countertransference and Seduction

As discussed in the section that lays out some of the central dynamics of borderline structure, issues surrounding the boundaries between self and others play out in many of their relationships and the relationship with the analyst is no different. What often arises from this confusion and diffusion of boundaries is a what the McWilliams (crediting Melanie Klein) refers to as projective identification, wherein the patient not only projects onto the analyst certain internal objects, but also elicits from the analyst a reaction as though the analyst where the object that the patient had projected onto them (McWilliams, 2011). One can also describe this dynamic qua transference: the borderline structured patient, as in the case of nearly all non-psychotic patients, positions the analyst in a particular way. However, unlike other non-psychotic patients, the borderline structured patient possesses a unique talent for eliciting from the other, in this case the analyst, a countertransference reaction that is a near-perfect reflection of that which the patient has transferred onto them. Ultimately, what seems to be at play in both descriptions of the dynamic is a certain kind of seduction of the analyst. Seduction here is defined quite liberally, to reference an attempt (conscious or unconscious) on the part of the patient to elicit from the analyst a countertransference reaction, not merely one that is romantic/erotic.

It is here, in this potential seduction that, again, the analyst can make a grievous mistake in the way in which they take up countertransference reaction. If the analyst allows herself to be seduced into becoming for the patient what the patient sees in her transference, then the boundary that separates the analyst from the patient and the analytic situation from any other relationship that the patient has or had dissolves. It is difficult to see, given what is understood about borderline structure, how an analyst sharing her emotional reaction to the patient’s behavior or disclosing our personal thoughts about the patient in the context of a so-called ‘interpretation’ that would not be akin to trying to put out a fire by tossing on an accelerant. For it is likely that, in revealing these thoughts and affects, the analyst has effectively given the response that the patient had so desperately wished to elicit. According to Campbell (1982), this sort of problem arises when a therapist, misunderstanding the borderline structure as well as the dynamics at play, attempts to conceive a real relationship in which the therapist shares personal thoughts and feelings. This dissolution of boundaries throws into question the very nature of the analytic situation for both the patient and the analyst: Are we friends who share a reciprocal relationship? Are we lovers, who each share our innermost selves with the other? Are we parent and child, where one cares for all the needs of the other?

In the case of Carmilla, the transference she experienced with Laura was romantic in nature, as was the countertransference she elicited. Like a borderline structured patient whose parent was sexually inappropriate or abusive and comes to see the analyst as a seductive or easily seduced parental figure might behave in a flirtatious or overtly sexual manner, Carmilla pursued Laura passionately. Laura described several scenes in which Carmilla, sometimes after one of her strange outbursts or fits of trembling would embrace her whispering tender apologies and ardent devotion. Sometimes, Laura noted “she would take my hand and hold it with a fond pressure, renewed again and again; blushing softly, gazing in my face with languid and burning eyes, breathing so fast that her rose and fell with the tumultuous respiration” (Le Fanu, 1871 p. 23). Though Laura was often troubled by these behaviors, even embarrassed she did nothing to stop them. She was taken in by the romance of it all, the beauty and strangeness of her companion, Carmilla’s over-powering passion, her own excitement of “a love growing into adoration” (Le Fanu, 1871 p. 22). Laura at times even fantasized that her companion, was, in fact, a young man, so taken with his love for and desire to be near her that he disguised himself as a girl in order to circumvent the mores of the day (Le Fanu, 1871).

It is easy to see how a similar dynamic could also arise in the context of an analysis. The intense emotions and constant drama of never knowing what will happen from one moment to the next in the course of analysis can have the all the flavor of an intoxicating romance. It is not the intention in this essay to demonize the analyst who experiences such countertransference reactions in the situation. Simply because an analyst experiences desire for their patient (whether it is romantic/erotic, parental, protective) does not mean she have allowed themselves to be seduced (as the word is used in the preceding paragraph). It is in her actions toward the patient, the interpretations she makes, what she chooses to disclose or express (McHenry, 1994). If the analyst has begun to act toward and speak to the patient as though the analyst were the patient’s lover, parent, friend, or any combination of the three, then she has allowed herself to be seduced.


Though the issue of how the analyst responds to the feelings they have for and thoughts they have about their patient is important in any analysis, it seems to present the greatest problem in working with borderline structured patients. In looking at the strange but compelling case of Carmilla, it is possible to illustrate all the signs of an analysis that has veered from its course and onto dangerous terrain: a relationship that begins disarmingly well and develops an intense and tumultuous quality wherein the caretaker’s interventions are guided primarily by a desire to continue the love and effusive flattery and avoid the animosity and burning ire of her charge. Amidst the drama and intensity, she allows herself to be seduced by the position that she has been placed in by a lovely and captivating young woman. The story can be viewed as a tragic one; for, in the end, only Carmilla’s death can end the suffering of both young girls. In case of an actual analysis, hopefully, the only tragic death suffered is that suffered by the analysis itself.

So what then, does all this mean for the analyst? What instruction can be gleaned in order to keep the analysis and patient alive when the analyst is faced with his own reaction to the borderline structured patient? Campbell offers up the suggestion that it is in the best interest of the borderline structured patient that the analyst not attempt to establish a real relationship with the patient but rather offer empathic understanding from a position of analytic neutrality (1982).


1. It is important for me to note that by using the story of a vampire as a case study, it is not my attention to demonize the borderline structured individual. Like the character in the story, the borderline individual is not inherently evil or malicious (or at least not more so than any human being). It is the way their structure positions both the real borderline structured patient and this fictional vampire in the world that affords them very little choice in how they relate to themselves and others. There is however, a compelling metonymy between the way in which this character and the borderline structured patient experience the world and others, as well as the way in which both need the other in a particular way that can be destructive and all-consuming.
2. The positioning of the analyst as the one who allows seduction to occur is intentional. Through their transference the patient is doing the work of analysis, and it is always the responsibility of the analyst to continue and support this work, in part, by providing a safe, non-reciprocal situation in which this work may take place.

Pages: 1 2

About the author:

at Duquesne University where she continues to pursue her interests in radical mental health, queer theory, the psychological and philosophical application of film and fiction, and to assemble found objects in cigar boxes.


Editor-in-chief: Kai Bekkeli
Assistant editor: Erica Freeman
Assistant editor: Monica Lawson
Assistant editor: Christopher Bailey
Advisory editor: Celeste Pietrusza
Faculty editor: Elizabeth Fein, PhD

Mike Fosnaught, IC-Dev, web designer

Back to Top