Des difficultés en théorie de l’esprit (TdE), une capacité à inférer chez autrui des états mentaux, intentions et émotions, ont été observées de façon répétée chez les personnes présentant une personnalité schizotypique ou atteintes de schizophrénie. Une tendance à effectuer des choix de façon hâtive dans des situations de raisonnement probabiliste a également été observée dans ces deux populations. Cette étude vise à examiner les performances à une tâche de TdE et à une tâche de raisonnement probabiliste chez des participants ayant un score élevé à une échelle d’anhédonie sociale (EAS), une des caractéristiques de la schizotypie, et chez des participants témoins. L’existence d’un lien entre score faible en TdE et choix sur la base de peu d’indices est également étudié. Conformément aux hypothèses, les résultats montrent que les participants du groupe expérimental ont des scores significativement inférieurs en TdE et demandent significativement moins d’indices pour effectuer un choix en situation probabiliste. En revanche, il n’existe pas de corrélation entre les performances de ces deux tâches. L’absence de corrélation peut être attribuée à une distance trop importante entre contextes expérimental et écologique pour la tâche de raisonnement probabiliste. L’intérêt de l’étude repose le choix de l’EAS comme critère de schizotypie en regard du type de tâche expérimentale proposé pour l’étude de la TdE.

Theory of mind (ToM) is defined as a capacity to infer mental states, intentions, and emotions in others. Two principal theories in the field of cognitive psychology have tried to explain mechanisms underlying this capacity. Theory-theory hypothesizes that people interpret cues from others in social interactions with a folk psychology composed of data about social human behavior. In contrast, Simulation-Theory proposes a capacity to take different perspectives and mentally simulate others’ behavior. As a result, one could guess a personʼs intention or emotion by comparison with his state of mind when he/she behaves in the same way.

Difficulties in ToM have been frequently observed in schizotypal subjects and subjects with schizophrenia. Some authors have proposed that this impairment could lead to persecution delusions or be linked with disorganized thought. A tendency to make choices with few cues in conditional situation has also been observed in both populations. When they are asked to make a decision about cues they can choose to see or not, schizotypal subjects and patients with schizophrenia tend to make up their mind after significantly less cues than control subjects. This tendency has been called “jump to conclusion”.

Our study tests the correlation between low performances at a ToM task and a tendency to jump to conclusion in conditional situation.

We tested this hypothesis with 25 participants scoring high on a social anhedonia scale (J Abnorm Psychol 85 (1976) 374–382), one of the main characteristics of schizotypy, and with 20 control participants. Participants were students with a mean age of 20. We included in the experimental group, subjects with a score on social anhedonia scale superior to 17 for females and 19 for males, and seven for control participants (modal score).

We used “jar and beads”, a conditional reasoning task. Two jars are presented to a participant: a white one containing 85% of white beads and 15% of black beads, and a black one filled with the opposite ratio. The participant has to decide from which jar comes successive beads shown to him. Dependent measure is the number of beads seen before decision. ToM task is an adaptation from (Schizophr Res 17 (1995) 5–13). Twenty short interactions between two characters are read to a participant. For example: John has a phone call with a friend for one hour. He says: “My mother ought to call me in a few minutes”. Question: What does John really mean? Cue: John adds: “I could call you tomorrow morning”. Question: What does John want to do?

Results show significantly lower performances at the ToM task in experimental than in control participants (52.36 (S.D., 6.73) vs. 59.05 (S.D., 1.60); t, 4.33; p<0.001; maximum possible, 60). The experimental group asked for significantly less cues to conclude in the conditional situation (2.22−S.D., 1.29). Mean number of beads asked for in the control group was 3.05 (S.D., 1.30) and t, 2.13; p<0.05. There was no correlation between performances at conditional reasoning task and ToM task. We observed this absence of correlation in all of the participants and in the experimental and control groups separately.

Absence of relationship between performances in both tasks may be attributed to a discrepancy between experimental and ecological contexts for conditional reasoning task. During interpersonal relationships, search for cues in order to make a decision about others’ intentions and mental states represents a real cost in terms of energy and time. These costs are absent in the “jar and beads situation”. Moreover, people with social anhedonia may attribute a special value of quickly understanding personal interactions. This conditional reasoning task does not imply this parameter.

Ecological decision in conditional reasoning tasks could be approached by adding a system of points to spend, asking for more cues, or to earn, finally finding the correct answer. Decision would then depend on the ratio between possible gains, by guessing or not the correct answer, and the cost of searching for more cues before making a decision.