Introduction:On one hand, there would be knowledge that is difficult to learn by oneself, that requires time, effort and high motivation (e.g., "academic" knowledge such as mathematics or grammar.). On the other hand, there would be knowledge so easy and quick to acquire that one could not even explain how to do it (e.g., to recognize kin faces, food, to speak a mother tongue). According to the evolutionary approach (Geary & Berch, 2016), the difference between these two types of knowledge not be a matter of complexity or even concreteness, but would consists in their adaptive utility and their time of acquisition during evolution. Primary knowledge would be knowledge for which our cognitive system has evolved making us motivated to acquire it quickly and effortlessly. Secondary knowledge would be knowledge for which our cognitive architecture would not have had enough time to evolve, therefore needing a lot of cognitive resources, time and motivation to be learned. In addition, several studies have shown that the way a logical problem is presented has a significant effect on task performance (Evans, 2005). This study proposes to test the characteristics of the two types of knowledge by using problem contents and additional extrinsic cognitive load. Compared with secondary knowledge, we expect primary knowledge to increase the participants’ efficiency in solving syllogisms and not to be impacted by an additional extrinsic cognitive load. We previously obtained results that confirm this hypothesis with conditionals (Lespiau & Tricot, 2018).Method:188 university students in France participated to this experiment. They had to solve syllogisms. We varied (i) the content of the problems (primary knowledge vs. secondary; e.g., food vs. grammar rules), (ii) the additional extrinsic cognitive load (via a Dot Memory Task with three modalities: low, medium and high cognitive load) and (iii) the type of syllogism (with or without belief bias, the former requiring more cognitive resources to be properly processed than the latter). Analyses assessed the influence of these variables on performance, problem solving speed and perceived cognitive load.Results and discussion:The results showed that participants were more effective (in terms of performance and speed of resolution) faced with primary knowledge contents compared to secondary knowledge contents. This positive effect of primary knowledge was particularly noticeable when problems required more cognitive resources to be solved (syllogisms with belief bias). However, the effect of the extrinsic additional cognitive load was the opposite of what was expected: the complexity of the Dot Memory Task did not influence the performance in secondary knowledge contents but that in primary knowledge contents. The higher the additional load was, the better the performance was, only for primary knowledge and especially for syllogisms with belief bias. This result is consistent with the evolutionary theory of knowledge since secondary knowledge would lead to cognitive conflict that would saturate cognitive resources and not allow participants to allocate sufficient resources to solve problems. On the contrary, primary knowledge that is supposed to consume very few cognitive resources would allow participants to process the additional load and to increase their performance despite this. This study also raises the hypothesis that a minimum cognitive load is necessary for participants to be invested in the task.