How Analysis Begins: The Preliminary Interviews
When Freud wrote “On Beginning the Treatment” in 1913, he described his practice of taking on patients “at first provisionally, for a period of one or two weeks,” which he called a “preliminary experiment” or “trial period.” His stated reason was pragmatic: “If one breaks off within this period one spares the patient the distressing impression of an attempted cure having failed.” Yet this seemingly practical arrangement points to something more fundamental—that analysis has a threshold, a passage that must be crossed, and that this crossing is neither automatic nor guaranteed.
Lacan termed this initial period the preliminary interviews. The expression itself indicates a structure: there is a before and an after, separated not by the mere passage of time but by a cut, a rupture that qualifies a change. This threshold is not the same as the entrance to the analyst’s office. It marks instead the entry into the new social bond of analysis. The question then becomes: what distinguishes these two moments, and what must occur for this passage to take place?
The Constitution of an Analytic Symptom
The person who first contacts an analyst typically arrives with what might be called a “collection of complaints”—insomnia, anxiety, failed relationships, a vague sense that “something is wrong.” These complaints, however diffuse or precisely articulated, do not yet constitute what psychoanalysis calls a symptom. For a symptom in the analytic sense is not simply suffering; it is suffering that has become a question, that points beyond itself to something the subject does not yet know they know.
The preliminary interviews serve first, then, to enable this transformation—from complaint to symptom, from a problem to be solved to an enigma to be articulated. This is not a matter of the analyst imposing a new vocabulary or reframing the patient’s concerns. Rather, through the very act of speaking to someone who listens in a particular way, something shifts. The complaint begins to resonate differently. Repetitions emerge. A pattern becomes audible that was not audible before.
What makes this shift possible? Precisely the analyst’s refusal to accept the demand in its raw form. When someone says “I want to be less anxious,” the analyst does not respond with a treatment plan, a diagnosis, or reassurance. The demand itself must be questioned—not dismissed, but interrogated. The analyst does not work for any norm, whether social or sexual. The analyst is employed in something else entirely: the elucidation of the torment of symptoms.
Lacan put it plainly: there is only one true demand that leads to the beginning of an analysis—the demand to be rid of a symptom. Everything else—the desire to “know oneself better,” to improve relationships, to become more functional—may be legitimate goals, but they are not demands for analysis as such. The preliminary interviews test whether such a demand can be formulated, whether the subject’s suffering has become, or can become, a question addressed to their own unconscious rather than to the Other’s knowledge.
The Emergence of Transference
“In the beginning of psychoanalysis is transference,” Lacan tells us. Yet transference itself requires time to develop, to manifest, to become workable. The preliminary interviews provide this time. What begins to emerge during these face-to-face meetings is a peculiar supposition: that the analyst knows something, that there is a knowledge about the subject’s suffering that exists somewhere, even if neither party can yet articulate what it is.
This is the essence of what Lacan calls “the subject supposed to know.” But here a crucial shift must occur for analysis to begin. Initially, the analysand supposes that the analyst knows—knows what their problem is, knows what they should do, knows the solution. The passage into analysis proper happens when this supposition shifts: when the subject supposed to know moves from the analyst as person to the unconscious itself. The analysand begins to address not the analyst, but their own unconscious through speech.
This shift is not always dramatic or easily detected. In both phases—preliminary and analytic—free association is at work. The analysand speaks, associations form, repetitions emerge. What marks the threshold is something more subtle: a change in the mode of address, a recognition that the speech is going somewhere other than toward the analyst’s response. The analyst’s role is to recognize when this shift has occurred and to mark it through what might be called an acceptance—the analyst accepts the analysand for analysis.
Clinical Structure and the Direction of Treatment
The preliminary interviews also allow the analyst to formulate a hypothesis about clinical structure—not to label the analysand, but to orient the direction of the work. This is not diagnosis in a medical sense. Psychoanalysis does not seek to “cure” the subject’s structure, which is incurable. What analysis can do is lead the subject to articulate a truth about their structure, to find within it a singular mode of satisfaction that is theirs alone.
The clinical structure matters because it determines what can and cannot be the reference point for treatment. For the neurotic, castration and the Name-of-the-Father organize the symptom; for the psychotic, these are foreclosed, requiring a different approach entirely. The analyst must know this—not to impose a treatment protocol, but to avoid ungoverned interventions that might foreclose the very possibility of the work.
The Cut into Analysis
Between the preliminary interviews and analysis proper comes a cut. This cut may be subtle or pronounced, but it marks a definite change. Often it coincides with the move from face-to-face meetings to the couch—the analysand reclines, the analyst sits out of view, and a new configuration of speech emerges. But the physical arrangement is less important than what it signifies: the analysand is no longer speaking to the analyst in the usual sense. They are speaking into a space where the unconscious can be heard.
What qualifies this cut? Alfredo Eidelsztein, in his book “The Graph of Desire,” emphasizes that analysis works step by step, from the simplest to the most complex. Yet the simplest—that someone speaks—is already complex enough. For speech in analysis is speech that does not know itself, speech that says more than it intends, speech that is determined by signifiers the subject did not choose. The preliminary interviews prepare the ground for this kind of speech to emerge. Analysis proper is where it unfolds.
The cut is also the analyst’s way of manifesting that they accept the analysand for analysis—that the threshold has been crossed, that a pact has been established. From this point forward, the fundamental rule applies: say whatever comes to mind. The impossibility of this rule, and the working through of this impossibility, will become the substance of the work.
Not Adaptation, Not Cure
It bears emphasizing what analysis is not seeking to accomplish. The principal position of the psychoanalyst is clear—the analyst does not work for any norm, whether social or sexual. Analysis is not about adjusting the analysand to social expectations, improving their functioning, or making them “mentally healthy” in a conventional sense.
Rather, the goal is finding a solution to suffering by elucidating the torment of symptoms. This solution is not given in advance. It cannot be prescribed or predicted. It emerges through the work itself—through the articulation of the symptom, the traversal of fantasy, the encounter with what Lacan calls the “free word” that the subject is most reluctant to speak.
The preliminary interviews, then, are where the possibility of this work is tested. Can a symptom be constituted? Can a demand for analysis emerge? Can transference develop such that the analysand will address their own unconscious rather than seeking the analyst’s knowledge? If so, the threshold can be crossed. If not, the work may remain preliminary—or it may be that analysis is not what is being asked for.
A Passage, Not a Program
There is no program for the preliminary interviews, no checklist to complete, no standard number of sessions. Their logic is not chronological but structural. What must emerge is a certain readiness—not a readiness of the ego, but a readiness of the symptom itself. For as Lacan emphasizes, analyzability is a function of the symptom, not of the subject.
This is why the preliminary interviews cannot be rushed or formalized. They unfold according to their own necessity. The analyst listens, questions, marks certain signifiers, allows transference to develop. The analysand speaks, elaborates, encounters the difficulty of saying what they mean. And at some point—recognizable in retrospect more than in the moment—the threshold is crossed. Analysis begins.