Conceptual research in psychoanalysis

By Dominique Scarfone

Psychoanalysis is not an experimental type of science, yet it is based on careful observation of psychic facts carried in the course of clinical practice. Psychoanalytic hypotheses are then empirically tested through the collaborative work of the analyst-patient team. The main difference with experimental research is probably that each analytic “team” is unique and intimately involved in the processes at study (observation is not independent of the observer), therefore not really allowing for reproducible results stricto sensu: what works with one analytic dyad does not necessarily so with another; secondly, the nature and duration of psychoanalytic treatment impedes the constitution of cohorts large enough for specific hypotheses to be prospectively tested. Moreover, confidentiality is not an external constraint but an intrinsic condition for the deployment of analytically usable material, which also limits reproducibility.

In spite of these limitations, there is no dearth of psychoanalytic theories and concepts, resulting in the often deplored “Babel” in the psychoanalytic domain. Jean Laplanche (1987), for one, suggested that the contemporary psychoanalytic conceptual field resembles the Ptolemaic astronomical system, where for every new observation a new “epicycle” is added to our theoretical model, rather than examining the eventual need for a “Copernican revolution” in our discipline. Laplanche went as far as proposing his own version of a Copernican turn in psychoanalysis, but we shall not discuss his proposal here. We will, instead, examine the method that he followed in his struggle against the spirit of Ptolemy in our midst, a method which inspires the sort of conceptual research that I try to carry.

Laplanche’s method of conceptual research

Laplanche’s research started with the delimitation of the psychoanalytic domain of study. For him it consists of four main fields of research: (1) Psychoanalytic treatment proper; (2) Extra-muros (a.k.a. “applied”) psychoanalysis; (3) History (mainly the history of psychoanalysis); (4) psychoanalytic theory.

The fourth element in this list is the one concerned with conceptual research proper. In retrospect, it is surprising that before Laplanche very few theoreticians of psychoanalysis, if any, had considered the possibility of applying the psychoanalytic method of investigation to Freud’s theory itself. Laplanche based his work in this field on the Freudian notion that the unconscious is a force affecting human thinking and discourse, eventually interfering with the establishment of adequate knowledge. Now, if this is true of any human thought, feeling and discourse, then it applies just as well to Freud’s own theory. It is not, however, a matter of psychoanalyzing Freud the man, but of considering theory itself as subjected to phenomena not unlike slips, negation, denial, rationalizing, reaction formation etc.

The reliability assumption

One important premise is that Laplanche’s method rests on the assumption that Freud’s thinking is reliable. That is: not that Freud was always right, but that he was relentlessly on the trail of the unconscious, tracking its various effects and manifestations. Such consistency in his effort made it so, according to Laplanche, that when Freud went astray in one part of his theorizing, this was somehow compensated with new thinking in another part. Both the parts may of course present problems, but the phenomenon as a whole is a good indicator of where lies the deeper problem in need of further research and reflection.

A good example of this is Freud’s introduction of a “death drive” in 1919, a controversial and problematic concept if there ever was one. For Laplanche, the seemingly sudden emergence of a totally new conception of the drives was in need of an explanation, to be found in what had been developed by Freud in the previous years. Indeed, based on the reliability assumption mentioned above, Laplanche posited that a change must have occurred somewhere in the preceding formulations, requiring that new concepts make up for the loss or the erasure entailed by that change.

Laplanche’s study has shown that Freud’s need for the concept of a death drive indeed resulted from the introduction of narcissism from 1910 on. Narcissism had entailed a sort of conceptual “taming” of the theretofore “daemonic” sexual drive, to the point where sexuality was now put on the same side as self-preservation instincts to the point where sexuality was now put on the same side as self-preservation instincts, under the larger heading of “life instincts”. Freud’s consistency in his positing the strangeness of the unconscious “thing” required, however, that the demonic aspect be relocated elsewhere in the theory. This was, according to Laplanche, the implicit reason for the introduction of a death drive which was now the element responsible for the turbulence of the drives that had been lost of view. An additional historical argument was brought by Laplanche in support of this view. It was discovered in the Minutes of the Vienna Society of the year 1910, where “life instincts” were mentioned nine years before Beyond the pleasure principle. The interesting fact is that they were then situated opposite to the sexual drives, suggesting that the placeholder for the daemonic sexual drives and the death drive was exactly the same.

From this example we can see that Laplanche’s method of theoretical/conceptual research requires (1) an extended knowledge of the Freudian theory; (2) an appreciation of the entire structure of the theory so as to detect the equilibria that prevail and which will eventually call for corrective measures whenever a new idea of concept disrupts those equilibria; (3) an epistemological study of the concepts themselves.

Working on post-freudian concepts and theories

Laplanche’s method can obviously be applied not just to Freud but also to any other “consistent” psychoanalytic theoretician. Most importantly for our present day situation, it can apply to the study of the impact that post-Freudian concepts have on the Freudian conceptual body taken as a whole. This puts on us an even greater pressure for first carefully examining the elaborate set of concepts that was Freud’s legacy in order to carefully monitor exactly what was added over the years by important post-Freudian authors and how their contributions relate to the original body. 

Here is an example. In the late 1990’s, while looking for a way of examining virtual reality from a psychoanalytic point of view, my student of the time Marie Leclaire and I were struck that in the vast majority of articles related to “reality testing”, this concept was taken for granted and could be summarized as meaning: “Reality testing is what is lacking in psychosis”, which is a rather poor definition for a concept of such importance. Going back to the Freudian texts we were struck by the different definitions Freud had given of “reality testing”, as highlighted by Laplanche & Pontalis in The Language of Psychoanalysis. So, we undertook the task of re-reading Freud on the subject and thereby were able to find that the concept went back to the very beginnings of psychoanalysis and that it entailed precise references to two different kinds of mnemonic traces as well as to a more detailed sequence of mechanisms than was usually assumed. We were thus able to document what we dubbed “actuality testing” as a first step towards “reality testing” (Leclaire & Scarfone, 2000). 

Recent work in the IPA’s Conceptual Integration Project Group (CIPG) has taught me many important things about how to examine concepts and their different meanings according to the various theoretical trends and “schools of thought” (Bohleber et al., 2013; 2015). The CIPG applied its method to single concepts, but I now think it useful to use it in comparing two or more concepts critically. Hinshelwood (2008) provided a recent example of such comparative study, though his specific method was not the same. His study examined splitting and repression and his results well illustrate, in my view, the interest of conducting comparative work. I believe that combining the methodology developed by the CIPG in the study of single concepts with a comparative method loosely inspired from Hinshelwood one can contribute to the effort of streamlining, so to speak, the conceptual “Babel” of psychoanalysis.

The parsimony principle

A deliberate strategic choice is implied here, which is to work from the premise that the Freudian conceptual body is still the fertile ground from which post-Freudian concepts grow. While these may employ a different vocabulary, it begs the question whether the introduction of a new term truly reflects the introduction of a new concept and is not a mere rebranding. A parsimony principle, attributable to William of Ockham (Fourteenth Century), but also to William James’ philosophical pragmatism, is necessary in this domain. The rule is that new concepts should be introduced only if they cover new ground, referring to some unique fact or process not already invoked by existing notions. In the physical sciences this rule is more easily obeyed than in our field since, in psychoanalysis, we lack the “crucial experiment” that demonstrates—or disconfirms—the uniqueness of the concept, let alone the very existence of the concept’s referent. The ease with which new ideas and expressions are brought to our attention in psychoanalysis certainly reflects the liveliness of the field, but it can also become problematic when expressions and ideas do not really add to our actual knowledge but simply reformulate older ideas in a new guise. Then again, this could still be a minor problem, were it not that such strategy—even when it is not intentional—results in a lack of genuine discussion among psychoanalysts and in the creation of ever more divergent theories and “schools” within psychoanalysis. 

One example could be the use of “dissociation” as a replacement for “repression” by a number of contemporary authors. One could argue that it is in fact quite difficult to set them apart, especially when both are looked at as a process. A close reading of Freud and Janet, for instance, shows how the two concepts, though different in some respect, are not really as different as modern theoreticians believe they are. If a well documented conceptual debate could be carried around these two concepts, it could possibly clarify the issue so that instead of a nominal preference for one of the other term, we could identify the true theoretical underpinnings of the distance between different “schools”.

Concepts and theories

Not surprisingly, the close study of any psychoanalytic concept immediately raises issues about other concepts belonging to the same conceptual cluster—think here, for example, of the cluster formed by primal and secondary repression, splitting (of the Ego, of the object), dissociation, suppression, denial, disavowal, negation etc.—so that conceptual research is, in a way, always examining the theory as a whole. Therefore, what Laplanche had postulated in Freud’s theorizing (that change in one part of the theory requires change in another part) remains true when we engage in research concerning contemporary issues. Thus, for instance, if there is indeed an—albeit partial—overlap between repression and dissociation, then we may wish to look for some underlying, more basic formulation that could account for the clinical phenomena encompassed by the two concepts. This, in turn, may question other large areas or clusters within the theory as a whole. For instance, one could ask how the more basic concept or “common denominator” behind dissociation and repression resonates with other common denominators in the remaining conceptual clusters, and so on. 

Lacking, as we saw, the “crucial experiment” and being observer-dependent, psychoanalytic practice cannot be expected to demonstrate the validity of its concepts in the way, for instance, the Large Hadron Collider recently confirmed the existence of the Higgs Boson. Our concepts and theories are supported by their potential refutability, their internal coherence, their compatibility with other concepts within existing conceptual clusters, their heuristic value and clinical usefulness. Indirect support can also be provided by their eventual resonance with concepts belonging to neighboring social and biological sciences. One must therefore identify concepts and theories that seem to rest on more shaky grounds and submit them to the test of seeing if and to what extent they meet the above mentioned criteria. In this way, conceptual research constantly “puts to the question” psychoanalytic theory as a whole and may thus ensure its vitality, its closer correspondence to the facts unearthed by psychoanalytic practice and its unique contribution, alongside other disciplines, to the task of better understanding the human condition.


Bohleber W, Fonagy P, Jimenez JP, Scarfone D, Varvin S and Zysman S (2013). Towards a better use of psychoanalytic concepts: A model illustrated using the concept of enactment. International Journal of Psycho-Analysis, 94, 501–530.

Bohleber W, Fonagy P, Jimenez JP, Scarfone D, Varvin S and Zysman S (2015). Unconscious phantasy and its conceptualizations : An attempt at conceptual integration. International Journal of Psycho-Analysis, 96, 3, 705-730.

Hinshelwood, R. D. (2008). Repression and splitting: Towards a method of conceptual comparison, International Journal Psycho-Analysis, 89, 503–521.

Laplanche J (1987). New Foundations for Psychoanalysis, New York : The Unconscious in Translation, 2016.

Leclaire M & Scarfone D (2000). Vers une conception unitaire de l’épreuve de réalité  Revue Française de Psychanalyse, Vol. 64 issue 3, p. 885-912.