The GeoCat 1.3, a simple tool for the measurement of Freudian primary and secondary process thinking

ABSTRACT

Linda A.W. Brakel et al. have developed a formal, non-verbal index of primary and secondary process mentation independent of the psychoanalytic clinical method, the GeoCat, and Ariane Bazan has made improvements to the original instrument, resulting into the GeoCat 1.3. The GeoCat is a forced-choice measure contrasting two types of similarity between geometrical figures. One of the two target figures consists of the same components as the master figure but in a different configuration; because this target figure is similar to the master figure in an “attributional” way, it is called the “ATT” figure. By drawing on associative operations to attain perceptual identity, ATT choices are proposed to index primary process mentation. The other target figure is made up of different components, but these are arranged in the same configuration; as this target figure is similar to the master figure in a configurational or “relational” way, it is called the “REL” figure. Configurational similarity is only accessible through spatiotemporal orientation, and therefore REL is thought to index secondary process mentation. Accumulating research on the validity of the GeoCat 1.3 shows good psychometric qualities, including internal consistency, inter-list variability, and both construct and convergent validity, with preliminary data showing high correlations with projective instruments. Even if the GeoCat does not deliver the rich clinical materials produced by projective tools, its easy inclusion in diagnostic assessment produces important, sometimes surprising, indications. As a research tool, it has proven pivotal in a variety of studies, from fundamental metapsychology and psychopathology to psychotherapy research.

KEYWORDS:

• GeoCat
• primary process
• secondary process
• Freud
• psychoanalysis
• measurement

Introduction

We introduce a simple tool for the measurement of primary and secondary process mentation, designed by Linda A.W. Brakel and colleagues in 2000, namely the GeoCat, and further developed by Ariane Bazan into its present version 1.3. The GeoCat is a formal, non-verbal index of primary and secondary process mentation independent of the psychoanalytic clinical method. After 20 years of its application, in this paper we provide a solid argument for its validity and discuss its utility for research.

The concepts of primary and secondary process are embedded in a complex metapsychological framework, which Freud developed to understand the constitution of the mental architecture in his Project for a Scientific Psychology (1895/1966). As such they are action pathways designed to deal with the “necessities of life” and distinguished categorically (by the mobilization, or its absence, of “indications of reality;” see Bazan, submitted). Facilitated by Freud’s neuronal neuronic descriptions, a number of neurophysiological explanatory frameworks propose coherent explanations for these processes, involving working memory and/or prefrontal cortex, the dorsal and ventral pathways, the efference copies and the default mode network (see Bazan, 2007a, 2007b; Carhart-Harris & Friston, 2010, 2012; for an overview, see Bazan, submitted). However, primary and secondary processes form cohesive mental phenomena with proper mental characteristics (see below), which are not individually related to neuroanatomic structures. Epistemologically, our endeavor is not only independent of all psychoanalytic concepts, but also to capture these mental processes, without prior consideration for their neurophysiological correlates, or, in other words, to measure the mental phenomena. However, we are then confronted with a new difficulty: indeed, measurement entails the selection of an operationalizable probe of the phenomenon, an (anecdotic) aspect of it, and not the whole mental richness of the concept. Warmth, for example, is a complex phenomenon with an immense theoretical elaboration, and its capacity to expand mercury is only a very anecdotic aspect; however, nobody doubts the effectiveness of the thermometer for its measurement, and nobody confuses warmth with the expansion of mercury.

Likewise, the GeoCat probes for primary and secondary process mentation by focusing on specific aspects of their mental phenomenology, i.e. how they manifest themselves in daily, conscious life: specifically, we propose to “pick up” primary process mentation by the proficiency to detect similarity based on common shared elements or attributes, and to register secondary process mentation by the proficiency to detect configurational similarity (see below). Indeed, we consider primary and secondary processes as two fundamental modes of mental organization, prior to specific thought contents, that permeate everyday functioning (see also Bazan, submitted). In this we follow Freud (1895/1966), who opposed the sharp dichotomy primary/unconscious and secondary/conscious processes and warns us “against possibly identifying primary processes with unconscious ones” (p. 340). In what follows, we first show how these proficiencies derive from their Freudian definitions and then how they correspond with the cognitive concepts of attributional and configurational (relational) similarity, the cognitive approach thereby providing a tool for their measurement.

In primary process mentation the “cathexis,” or activation quantity, flows along the most facilitated pathways and one neurone complex discharges its cathexis into other neurone complexes. This sequential transfer results in “analogous states” (Freud, 1895/1966, p. 325) and constitutes an associative dynamic – a most essential characteristic of the primary process. Indeed, Freud (1895/1966, p. 338; 1900/1955, p. 597; 1901/1978, p. 278) describes the clinical manifestation of primary process respectively as “a compulsion to associate,” “superficial associations” and “external associations.” In other words, the primary process assembles representations on the basis of shared elements, even if these are non-essential or coincidental elements. Primary process links would for instance, assemble all cylindrical objects, from pipes to pens, or all blue things, from dresses to ducks, due to only sharing that one characteristic. Accordingly, Rapaport (1951) claims: “Where the primary process … holds sway … everything belongs with everything that shares an attribute of it” (p. 708). Holt (1967) also speaks about an association on the basis of “non-essential” features (p. 354). The attribute, in a primary process mode, may come to represent metonymically the whole thing; the principle is called pars pro toto, the “part for the whole.” For example, in his discussion of “joke work” as analogous to dream work, Freud (1905/1960) identifies primary process mentation as follows: “displacement, faulty reasoning, absurdity, indirect representation, representation by the opposite,” and he elaborates “indirect representation – the replacement of a dream-thought by an allusion, by something small” (pp. 88–89). With waxed, long upward curled mustaches alone, we allude to Dali; with a bowler hat and a bamboo cane, we indirectly represent Charlie Chaplin: these are manifestations of primary process thinking. This identification based on common attributes is the perceptual identity (Freud, 1900/1955) which the primary process pursues. Clearly, this is a frequent operation in common conscious thinking. This is the case, for example, for identifying objects in our visual environment of which we only see a (small) piece: by seeing its tail, we know the cat is behind the chair, by seeing only a corner of the table, we “see” the whole table, etc.

This associative principle of the primary process applies both to conscious and unconscious processes; however, in unconscious processes, such as in dreams, cathexes transfer quite completely from one representation to the other. Next to simple association this gives easily rise to displacement and condensation:

By the process of displacement one idea may surrender to another its whole quota of cathexis; by the process of condensation, it may appropriate the whole cathexis of several other ideas. I have proposed to regard these two processes as distinguishing marks of the so-called primary psychical process. (Freud, 1915, pp. 186–187; italics added)

Laplanche and Pontalis (1967) describe displacement as the detachment of the intensity of a representation which is passed over to an associated representation which was originally not very intense, or even insignificant, but which now hides the first. Bergeret (1972/2004) describes displacement as “the flow, the slip of cathexis along an associative path, linking diverse representations, thereby putting one representation in place of the other” and explains that this is a means of defense of the system Unconscious (Ucs) to try to pass the censorship (p. 63, italics added).

As the primary process is associative, the thinking slides over “analogous states,” as it were, on a horizontal plane. A “third powerful factor” (Freud, 1895/1966, p. 323) will develop with the maturation of the ego, but without this third point, primary process thinking cannot step outside of that plane and consequently, there is no vantage point from which to consider the thoughts from a distance. Therefore, in predominantly primary process thinking, representations are difficult to order reciprocally, even consciously. As a result, there is confusion, both for the hierarchical (e.g. as concerns their importance) and spatiotemporal arrangement of representations. Again, in the Ucs, this principle is more marked: Freud (1915) calls this the “exemption from mutual contradiction” (pp. 186–187): “Thoughts which are mutually contradictory make no attempt to do away with each other, but persist side by side” (Freud, 1900/1955, p. 596) and

When two wishful impulses whose aims must appear to us incompatible become simultaneously active, the two impulses do not diminish each other or cancel each other out, but combine to form an intermediate aim, a compromise. There are in this system no negation, no doubt, no degrees of certainty. (Freud, 1915, pp. 186–187)

The categorical difference applied by the secondary process mentation upon the primary processes is the introduction of inhibition. Freud (1895/1966) is very clear: “those processes which … represent a moderation of the [primary processes] foregoing, are described as psychical secondary processes” and will remain so consistently afterwards, e.g. “I propose to describe the psychical process … which results from the inhibition imposed by the second system [the memory system or Ego] as the ‘secondary process’” (pp. 326–327; italics added) and “it is only during the course of life that the secondary processes unfold, and come to inhibit and overlay the primary ones” (Freud, 1900/1955, p. 601, 603). Saraga and Gasser (2005) qualify inhibition as the “essence itself of the secondary process” (p. 111). This inhibition first comes with the developmental modifications resulting from experience (history), such that a ramified system (the ego) “mechanically” refrains the progress of the stimulus; moreover, in this ego a store of excitations or cathexes will be accumulated, constituting, as already indicated, a “third powerful factor” (Freud, 1895/1966, p. 323).

First, this allows for goal-directed action: Freud (1895/1966) explains that it is this cathexis, accumulated in the ego, which is in a position to counteract – or inhibit – the facilitated pathways. Indeed, these facilitated neurones put “difficulties in its [the secondary process] path towards establishing ‘thought identity’” (Freud, 1900/1955, p. 602), an identity concerning the purposive idea (see above). To establish thought identity, this thinking “must concern itself with the connecting paths between ideas, without being led astray by the intensities of those ideas” (Freud, 1900/1955, p. 602). In other words, the cathected goal organizes the secondary process.

Second, this allows for spatiotemporal arrangement. Primary process mentation indeed is characterized by spatiotemporal confusion. For example, Anna Freud (1936) indicates: “[when] the so-called ‘primary process’ prevails, there is no synthesis of ideas, affects are liable to displacement, opposites are not mutually exclusive and may even coincide and condensation occurs as a matter of course” (p. 7, italics added). Lacan (1964/1979 [1973]), concerning the primary process, proposes that “what happens there is inaccessible to contradiction, to spatio-temporal location and also to the function of time” (p. 31, italics added) and Kapsambelis (2005) describes the primary process as a mental process where “no obstacle, no logical or deductive link, no spatial or temporal constraints, disrupt the rapid flow from one representation to another” (p. 64, italics added). Brakel (2009) characterizes the primary processes as those taking place in a “tenseless and unexamined present” (p. 114). Finally, Bazan (2007a) underscores the positive possibility of the secondary process for spatiotemporal (and grammatical) distinctions.

Cognitive research in the 1980–90s has identified two different categorization principles: categorization based on attributes and categorization based on relationships (Medin et al., 1990, 1993; Smith & Medin, 1981; see Figure 1). The notion of attribute refers to any singular property constituting a stimulus, while the notion of relationship refers to the links between the different attributes of an object. In the case of a human face, eye color, nose size, mouth shape, skin color, and many other criteria are the attributes of this face. On the other hand, notions such as the distance between the eyes, the color of an eye compared to the other, the size of the nose compared to the mouth, in turn, involve the concept of relationship. In attributional categorizations, exemplars are regarded as similar if particular features match, even if relationships among the features are quite different: stimuli are classified on the basis of attributes or superficial features. As attributional similarity judgment is based on feature-by-feature associations and on attribute-based perceptual identity, it reflects and can thereby index primary process mentation. On the other hand, relational similarity judgment builds on similarity of relations between parts. The specific type of relational similarity, which is used as a probe in the GeoCat, is configurational similarity. Configurational similarity requires not only an insight in mutual relations of the attributes but, crucially, access to spatial judgment. As explained, only the secondary process mentation enables spatial and temporal distinctions. For this reason, the recognition of configurational similarity can indicate secondary process mentation.

Figure 1. Example of a triad of one item of the GeoCat consisting of a master figure and two target figures. The ATT-target (left) consists of the same components as the master figure but in a different configuration and is thought to probe for primary process mentation; the REL-target (right) is made up of different components, but these are arranged in the same total configuration and is thought to probe for secondary process mentation.

In summary, the GeoCat, which offers the participant a forced-choice similarity judgement between geometrical figures, proposes that choices based on attributional similarity (termed ATT-choices) require associative a-rational connections devoid of perspective-taking and therefore index primary process mentation; while choices based on configurational similarity (which is a type of relational similarity and are therefore called REL-choices) require the mobilization of a vantage point allowing for perspective as well as the ability to refrain from associative thinking, and therefore index secondary process mentation. The take we propose on primary and secondary process mentation has many parallels with Kahneman’s distinction between System1 and System2 (S1 and S2; Brakel & Shevrin, 2003; Holt, 2009, pp. 26–27¹). Indeed, S1 “operates automatically and quickly, with little or no effort and no sense of voluntary control,” like an “associative machinery” or “a machine for jumping to conclusions” and “neglects ambiguity and suppresses doubt,” while S2 can “construct thoughts in an orderly series of steps” and its operations require attention, effort and control (Kahneman, 2011, pp. 22, 52, 104, 23, 24). However, Kahneman’s view, as other general dual process theories, suggests a sharp hierarchical distinction between instinctual/emotional versus thought/rational processes (e.g. Matte Blanco, 1975; Modell, 2014; Panksepp, 1998; Russ, 2020). This differs from our approach. Indeed, we consider both primary and secondary processes first as action and/or thought pathways, even if indeed high emotion more readily induces primary process mentation. We think the action/thought-approach is the truest to the Freudian project. As important, it makes the most sense in modern neurosciences (for review, see Bazan, submitted), and moreover, it allows for the proposed index of its measurement – this is, outside of psychoanalytic theory and clinical psychoanalysis. Indeed, it is important to underscore the GeoCat’s epistemological independence, at it utilizes only cognitive concepts (of similarity), with no a-priori psychoanalytic relevance. The research questions underlying this review paper concerning the GeoCat then become: does this (cognitive) take on primary and secondary process mentation hold when confronted with empirical evidence, and additionally: what further evidence can we gather to substantiate its validity?

Materials and methods

The original paper-and-pen GeoCat consists of a single sheet with six squares, within which each square has a triad of geometrical shapes. For each triad, the top figure is hereafter referred to as the “master figure,” and the figures below are called “target figures.” One of the two target figures consists of the same components as the master figure but in a different configuration; this target figure is similar to the master figure in an attributional way and is hence abbreviated ATT. The other target figure is made up of different components, but these are arranged in the same total configuration; this target figure is similar to the master in a configurational way, which is a specific type of relational similarity and hence this target is abbreviated (for historical reasons) REL. The experimental design, in which participants are asked to make similarity judgments, is predicated on the assumption that a target choice based on its attributional similarity with the master figure reflects a primary process-based assessment, while a target choice based on its configurational similarity with the master figure reflects a secondary process-based assessment.

There are four series: 1A and 1B, and 2A and 2B are identical except that the two lower target figures in the triads are reversed (left-right). This controls the possible effects of lateralization in target choice. For the triads of the 1A and 2B series, the two target figures are identical; however, as the master figure is different, whenever a particular figure represents the ATT target choice in 1A, it is the REL target in 2B, and vice versa. The same reversals hold for series 1B and 2A. This controls the possible effects of preference for the figure itself in the target choice. In quantitative research it is therefore important to alternate or randomize the forms of the 4 series over the participants. The forms of the GeoCat 1.3 are given in the Appendix; they may be used without copyright payment but should be referred to as “GeoCat 1.3” with due reference to this paper as well as to the original GeoCat publication (Brakel et al., 2000).² When using an electronic version, the stimuli of one series are presented in a random order, and the same copyright rules pertain. The instruction used is:

There are six items on this sheet. Each item has a top central figure and two figures below. On each of the six items: Look at the top central figure; decide which of the two figures below is most similar to the top central figure. Circle your answer to each. There are no right or wrong answers. It’s your choice that counts.

If the subject indicates that s/he understands that there are two types of systems possible and asks according to which (categorization) system s/he should give an answer, the researcher simply repeats: “There are no right or wrong answers. What we are interested in, is your choice.”

The child version (2.5–6 years) is in all respects similar to the standard version, except that the six boxes of triads are presented one by one on an A4 format sheet of paper. The six stimuli are shown to the child, sheet by sheet, following the order of 1–6. The administrator first points to the master figure, temporarily masking the two target figures to effectively guide the attention of the child to the top figure above and says: “Look at this figure.” Then the researcher unmasks the two target figures below so that the child has an overview of the three figures together and asks: “Show me now which of these two figures is most like the top figure.” It is possible to repeat the instruction to the child for each presentation of a new triad. For the very young (2–3 years), we used the instruction “which is (most) the same?” As even the word “most” can lead to confusion, it can be left out. While the child points to the figure of his/her choice, the researcher circles this choice on the standard version corresponding to the child version. With children, we standardly do a second round, where we give them the choice they first made and ask them why they made this choice: we note the child’s response adjacent to the corresponding item on the standard version (e.g. “it’s like a crocodile”). With 2.5 year-olds, we have sometimes used a color-pretest where a master disk of blue with two target disks (one in red and one in blue) was presented together with the question “which is the same?,” in order to be sure that the child understands the instruction. However, this pretest also sometimes created confusion as the actual test did not concern colors.

Results

Internal consistency of the GeoCat is very good. Vanheule et al. (2011), using GeoCat 1.0, found Cronbach’s α’s of .88 for Series 1 and .96 for Series 2. The first author, using GeoCat 1.3, found an overall α = .94 in an adult population (N = 1,233; μ = 2.6 ± .07; median = 2). This indicates that the items from both series measure a single, unidimensional latent construct. Also, the results did not differ according to the series (1A, 1B, 2A or 2B; χ²= .828; p = .843).

Evidence for the construct validity of the GeoCat comes mainly from the observation that a majority of ATT has been found in all situations where primary process mentation is supposed to dominate, according to Freudian metapsychology. First, Freud (1895/1966) had proposed that primary process mentation is prevalent in unconscious mental life. In a subliminal priming paradigm, participants were presented with GeoCat 1.0 stimuli both unconsciously (subliminally) and consciously (supraliminally) (Brakel et al., 2000) and indeed, in strict subliminal conditions significantly more attributional than relational judgments were made (p < 0.008, unilateral testing).

As concerns young age, Freud (1900/1955) observed that

the primary processes are present in the apparatus from the beginning, while the secondary processes only take shape gradually during the course of life, inhibiting and overlaying the primary, whilst gaining complete control over them perhaps only in the prime of life. (p. 603)

Indeed, in GeoCat 1.0 research with participants ranging in age from 3–90+ years, it is observed that preschoolers till the age of 5 make significantly more ATT choices, and this changes sharply with a real shift to predominantly REL at 6. This predominance of REL then remains stable throughout life (Brakel et al., 2002). In an ongoing study with GeoCat 1.3 in a French-speaking Belgian population (in preparation), it was found that the mean ATT gradually increased between 2.5 and 5 years (with population means or μ’s from 3.6 to 4.5 ATT choices on a total of 6 possible ATTs) and then suddenly switches back to 2.9 ATT choices at 6 years, to stabilize around 2.1 in the further years up to age 12.

Third, Freud (1895/1966) also proposed that with high emotion, (motor) pathways are readily taken which are normally avoided such that “the affective process approximates to the uninhibited primary process” (p. 357). Primary process mentation is thus thought to dominate in highly emotional anxious states, anxiety being related to unconscious conflict according to Freud (1926). Indeed, in a naturalistic study, 120 patients (19–69 years) in medical center waiting rooms rated how anxious they felt on a 10-point scale and then completed the GeoCat 1.0 (Brakel & Shevrin, 2005). Those who reported any anxiety at all showed a significant shift towards ATT over those participants who rated themselves as calm (41.7% versus 25.5%; p < .02), which was interpreted as a shift in favor of primary process mentation. In a later replication study (Lamot, 2014), including 78 patients in a medical waiting room of an AIDS testing facility (mean age = 30,6; 35% women) and 72 non-patient controls (mean age = 31,6; 38% women), ATT correlated significantly with a self-measure of anxiety on a visual analogue scale, comparable to the Brakel and Shevrin study (r = .17; p = .042) but, remarkably, did not correlate with the State-Trait Anxiety Inventory scores (STAI). In a population of patients diagnosed with an anxiety disorder, De Gélas (2015) shows a high prevalence of ATT in the clinical compared to a control population (μ’s of 2.8 versus 1.7 on 6; p < .01). Furthermore, ATT in the clinical population correlates both with trait anxiety (STAI-Y2; r = .37; p < .005) and, to a lesser extent, with state anxiety (STAI-Y1; r = .21; p = .11). These last results seem to show that anxiety as a stable personality trait structurally influences mental processing, more than state anxiety due a stressful situation does. In a general population (N = 58; 52% women; 20–29 years), Renier (2012) has shown good correlations between ATT and state anxiety (STAI-Y1; r = .44; p = .000), trait anxiety (STAI-Y2; r = .23; p = .049), and VAS self-rated anxiety (r = .47; p = .000). Such clearcut results have not been replicated since. We think that they were due to the very homogenous population with low self-rated state and trait anxiety, a low social desirability score, and finally low mean ATT (μ = 1.5 out of 6). Remarkably, this has also been the only study with a correlation between ATT and a social desirability measure (the Marlowe Crowne Social Desirability Scale; r = .38; p = .001).

Fourth, primary process mentation is thought to dominate in acute states of psychosis. Indeed, in the Freudian model of psychosis, psychodynamic processes typical for the unconscious are present in conscious mental life:

As regards the relation of the two psychical systems [the conscious and the unconscious], all observers have been struck by the fact that in schizophrenia a great deal is expressed as being conscious which in the transference neuroses can only be shown to be present in the Ucs by psycho-analysis. (Freud, 1915, p. 197)

Fenichel (1945) agrees: “In non-psychotic persons, this mode of thinking is still effective in the unconscious. Therefore, the impression arises that in schizophrenia the unconscious has become conscious” and he adds:

“Because the primary process and the archaic ways of thinking have come to the fore again, schizophrenics are not estranged by these mechanisms any more” (p. 422; italics added). In particular, positive psychotic symptoms, such as hallucinations, perceptual distortions and delusions, function along primary process principles (Freud, 1895/1966, 1900/1955). Indeed, Bazan et al. (2013) show that psychotic patients (N = 72) tend to have more ATT choices than non-psychotic patient (N = 55; μ’s on 6 of 4.0 versus 3.3; p = .06, one-tailed) and patients with acute psychotic symptoms (N = 36) show significantly more ATT responses than patients without acute psychotic symptoms (N = 36; μ’s on 6 of 4.3 versus 3.4; p = .019, one-tailed). Further, GeoCat 1.3-results of Parrent (2011), De Smedt (2014), Goddefroy (2014) and Delannoy (2017) show that the more the delusion of the psychotic subject is articulated and elaborated, the less the thought processes operate in a primary process mode: a substantial negative correlation between numbers of ATT-choices and articulateness of speech³ was found (N = 8, r = −.87, p < .005; Parrent, 2011 and N = 13, r = −.61, p = .013; Delannoy, 2017).

Finally, Bydlowski (2001) suggested that pregnant women show a mental state qualified as “psychic transparency,” characterized by a dimming of the usual defenses and an overflow of reminiscence, fantasies and infantile representations, i.e. a predominance of the primary process. Indeed, ATT in pregnant women between the 6th and the last month (N = 118) was substantially higher than in non-pregnant control women (N = 81; μ’s on 6 of 3.5 versus 2.1), and also higher than in future fathers (N = 78; μ = 2.4; p = .001) – which did not differ significantly from control men (N = 40; μ = 1.8; Bazan et al., 2019). The hypothesis here is that the mere event of having another being growing inside one’s own body, would, by requiring huge mobilization of mental energy, deplete the mental resources required to maintain the usual defenses and as a consequence, enhancing primary process mentation, resulting in a so-called psychic transparency state.

Another piece of evidence for the construct validity comes from the factor analysis made with the GeoCat 1.0 (Vanheule et al., 2011), which showed with the Confirmatory Factor Analyses that the assumed model with one latent factor reflecting ATT and REL similarity fits particularly well to the Series 1A & 1B response pattern and is acceptable for the Series 2A & 2B response pattern. The results show that all items have substantial factor loadings and thus are good indicators of REL and ATT categorization processes, but the amount of variance explained in item 6 of both series is substantially lower than in the other items. Accordingly, these items were adapted in GeoCat version 1.3.

Good convergent validity has also been shown between the GeoCat 1.0 and a scalar version of the GeoCat, the GeoScalar (Vanheule et al., 2011; N = 399; 87% women; mean age of 19.4). The scalar version of the GeoCat presents a master figure together with one of both target figures, ATT or REL, on a single sheet and the participant is asked to judge the similarity between two items on a Visual Analogue Scale. The GeoScalar is not a forced-choice instrument: it does not require to choose between ATT or REL. The GeoCat ATT correlates .36 (p < .01) with the GeoScalar ATT and −.34 (p < .01) with the GeoScalar REL. Even with the GeoScalar, the REL similarity score is higher than the ATT score (μ on 5 of 2.4 versus 2.2; p < .01). Interestingly, the GeoScalar ATT did not correlate (r = −.01; p = .80) with the GeoScalar REL. Most GeoCat items explain significant amounts of variance in the GeoScalar subscale score, which indicates that participants who chose ATT more frequently on the GeoCat also have higher GeoScalar ATT scores and lower GeoScalar REL scores and vice versa. The 4 items in the versions of GeoCat 1.0 for which this did not hold true were changed accordingly in GeoCat 1.3. Furthermore, in this same study it was found that participants who completed the GeoCat first had higher GeoScalar REL scores than participants who first completed the GeoScalar (B = −.65, p < .01); however, there was no influence on the GeoScalar ATT scores. There was also no effect of prior completion of the GeoScalar on subsequent GeoCat scores.

Preliminary data on the convergent validity of the GeoCat 1.3 with projective tools is promising. First, the TAT, consisting of 18 illustrated plates, allows for the analysis of the different “speech processes.” Among these, the E-Series measures the emergence of primary processes (Brelet-Foulard & Chabert, 1990/2003), including alteration of perception, massive projections, identity disorganization, and speech alteration. An exploratory study (Parrent, 2011) including 8 psychotic subjects, a woman and 7 men, aged between 19 and 43 years, hospitalized by therapeutic injunction within the same closed psychiatric department and all going through an acute psychotic episode, showed a very strong positive correlation between the ATT and the proportion of primary processes in TAT-protocols (r = .95; p < .005; see also De Gélas et al., 2017). Second, according to Chabert (2012), primary process mentation can also be revealed on the Rorschach through several elements, including outside-the-box demonstrations, but also extreme vigilance, reluctance, or mistrust of the subject when faced with the task being asked. The subject avoids this by engaging in remarks that show a persecutory experience. Richelle et al. (2009) also propose that an emergence of primary processes, sometimes a possible persecutory experience, is likely to appear in non-responses, a dissociation of language, or even in interpretative elements. On the basis of these theoretic elements, Delannoy (2017) constructed an index for the measurement of primary process mentation in Rorschach protocols. In a population of 13 adults, composed of 10 psychotic subjects (4 in a residential hospital, 3 in day care, and 3 in a medico-psychological center) and 3 non-psychotic patients of a residential hospital, ATT and the Rorschach primary process-score correlated strongly positively (r = .72; p = .003). These very high correlations are in need for replication but they do indicate that, despite being two tests of a very different nature – projective for one, perceptive for the other – they seem to measure the same construct.

Discussion

All the psychometric data tested to date – internal consistency, inter-series variability, construct validity and convergent validity with a similar measure (GeoScalar) as well as with totally different measures (TAT and Rorschach) for the same construct – are good to very good. In particular, even without probing for the emotional or “instinctual” character claimed by several authors as essential to primary process mentation (see Introduction), we find coherent increases in primary process mentation with the exclusively cognition-based GeoCat in those clinical situations where this increase was predicted by Freud: unconscious processing, young children, clinical anxiety, and acute psychosis, as well as a condition not predicted by Freud but often correlated with high emotion, late pregnancy. Therefore, despite its simple format, we do think that the GeoCat validly measures the current equilibrium of primary and secondary process mentation and is an especially good measure for the primary process score, which might be a more or less stable personality trait. Equally important, with the GeoCat a number of seemingly disparate clinical conditions and situations, which are in themselves not obviously related, gain a theoretical coherence in a parsimonious way: they are all manifestations of predominant primary process mentation.

The GeoCat has undeniable advantages over other measures. First, it is extremely simple and rapid, which makes it easy to use and to score, and there is no need for a specific training as there is for the Rorschach and the TAT. Second, it is an answer to the Grünbaum (1984) circularity criticism in that it has no embedded psychoanalytic concepts, and since it requires no background knowledge, researchers outside of psychoanalysis can use it. Third, as it is non-linguistic it allows for measures in all cultures, in small children, and even in animals (e.g. Garlick et al., 2011), opening the way for evolutionary considerations (see Brakel, 2017; Cutler & Brakel, 2014). Fourth, it has no face validity. Face validity indicates the degree to which the participant has the impression that they understand what an instrument measures; as it refers to what the tool seems to measure, it is not a validity in the technical sense of the term. We believe it is quite impossible for the participant to guess the underlying theoretical concept of the GeoCat. This is a definite advantage, since the results cannot be biased by a strategy or by social desirability. Finally, its slightly playful character with forms and shapes is not daunting for participants, nor has it any suggestive effect considering pathology, defaults or abnormalities. Therefore, we consider it also a very ethical tool.

Of course, the GeoCat also has limitations. First, it doesn’t deliver the rich clinical material which is gathered by a TAT or a Rorschach. For this reason, it is very limited in its use as a singular tool in a diagnostic assessment. A GeoCat result should never be interpreted in isolation, but can deliver supplementary information in a complete assessment. For example, in two patients with so-called “aphantasia” (Collignon et al., 2022), there was an a priori assumption that given their lack of visual imagery there would be less primary process mentation. However, the GeoCat showed ATT results above the mean, and this surprising score helped orient the diagnostic analysis into another direction, which eventually proved coherent. In two other studies in a general population, one concerning obsessional tendencies (Corrales-Cricboom, 2021) and one concerning eating disorders-tendencies (Guillen & Vidaluc, 2020), it was surprisingly found that the higher the intensity of the symptoms, the higher the ATT-scores, though the general correlations in these non-clinical populations were modest. The GeoCat might thus prove a revealing supplementary tool in a general diagnostic assessment.

Third and importantly, the fact that it is a forced-choice tool imposes limits: specifically, information about the secondary process might be missed. Indeed, the results of Vanheule et al. (2011) confirm that primary and secondary processes function independently. They also suggest that the primary process functioning might be quite stable, functioning as a structural characteristic or a trait, while the secondary process functions in a manner that is more contingent and sensitive to contextual elements. For this reason, we think that the GeoCat is primarily a measure for primary process mentation. Fourth, and not unrelated to the previous, the fact that it is a non-linguistic tool might also be a disadvantage for secondary process indexation. As said, the essence of the secondary process is inhibition, and it is probably less evident (and maybe altogether impossible) to inhibit geometrical figures. Previously, Bazan (2007a, 2012) and Bazan and Snodgrass (2012) have proposed that repression is a special instance of secondary process – that of the inhibition of linguistic fragments of particular emotional intensity. We propose that the GeoCat is not sensitive to that linguistic inhibition dimension of the secondary process, and therefore doesn’t measure it. For example, in a recent study in which the occurrence of laboratory-induced slips (Motley & Baars, 1976) was explained in terms of primary process mentation and secondary process inhibition, the ATT scores were used for the primary process parameter. This parameter formed a baseline in participants with both low and high defensiveness scores, but for highly defensives another measure, specifically tailored for secondary process linguistic inhibition – namely the number of non-related word choices N in a forced word choice task between N and a phonological inverse P of the master word – explained another portion of the variance. In these participants, ATT and N/PN-scores together explained 30% of the variance of the occurrence of parapraxes (Thieffry et al., submitted).

As a final limitation, the GeoCat-score is quite sensitive to order effects: interventions which induce “secondarization” are thought to skew the GeoCat scores towards REL. For these reasons, we suggest including the GeoCat always as the first measure in a protocol. Relatedly, in other unpublished research we have noticed that when participants were technically able to go back to previous items to change them in online-protocols (e.g. when they discovered the “REL-principle”), this biased the general results. Therefore, we suggest that in electronic or online versions, it should be made impossible for participants to scroll back to previous items.

As concerns the limitations pertaining to the present review, some psychometric data are clearly missing, such as, principally, a test-retest reliability score and a confirmatory factor analysis for GeoCat version 1.3. Also, part of the presented data come from unpublished Master’s dissertations, and therefore need confirmation in peer-reviewed published formats – this is especially the case for the convergent validity data with the TAT and the Rorschach.

Concluding considerations

Since its first design in 2000 by Brakel et al., the GeoCat has accrued modest but steady success, with a growing demand for the use of the tool from outside its original crucible (e.g. Finn et al., 2017). Research in psychoanalysis is mostly focused on the singularity of the clinical case, which explains why tools like the Rorschach and the TAT, delivering rich detailed materials, are often preferred to bare GeoCat numerical results. However, even in clinical diagnostic assessments, GeoCat results, which are very easy to obtain (with no cost in time or effort from the participant), can sometimes subvert clinical preconceptions. Moreover, the potential of the GeoCat in research is exciting. It has been very interesting, for example, to show that psychotic patients, who could develop an articulate delusional system, functioned less in primary process mode, giving empirical strength to the therapeutic value of an elaborate delusion – be this delusion elaborated as a form of self-treatment by the patient himself, or, sometimes, within the course of therapy. These kind of results serve to substantiate clinical, and even therapeutic claims. Then, too, the possibility of assessing primary process predominance in non-human animal species would allow a welcome expansion of psychoanalytic-based concepts to the wider intellectual world. The purpose of this review was to show that for all these important research and clinical purposes, and probably beyond, the GeoCat 1.3 is a very easy, validated tool, which escapes the Grünbaum criticism, and which is moreover ethical in its use towards participants.

Acknowledgements

A. Bazan thanks the International Psychoanalytic Association for financial help with the publication fee. The author also wants to thank Mr. Bob Berry for his unwavering support over the years. Ariane Bazan is also profoundly indebted to Howard Shevrin for his invaluable mentorship.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research received no funds.

Notes

1 This is not amazing: Kahneman spent the summer of 1960 as a student of David Rapaport at Austen Riggs, a psychoanalytic hospital, and Rapaport (1951; see also Gill, 1967) himself was a well-known Freud scholar, especially expert on Freud’s metapsychology, and elaborately studied the so-called Freudian primary and secondary processes.

2 They are also free for translation to other languages, but the translated version must also come without copyright (e.g., for the translator) and with the same references (this paper; Brakel et al., 2000).

3 The articulateness was measured by indices formally evaluating the participant’s speech in his responses upon the “Questionnaire of the Expert” in which the participant, taken as the expert, is consulted upon various societal and existential topics, e.g. How should we educate children? The underlying idea is that a psychotic delusion (but also a neurotic mythical story) is an answer to the basic existential questions of life (where do we come from?, what will be left of us after us?, how should we live? etc.). In the “Questionnaire of the Expert” we submit the participant to questions inspired by these existential themes; the idea, then, is not to measure the content but the formal structure of the speech in response to these questions.