The question of associativity, which is so fundamental in standard analysis, presents itself differently with patients functioning in a psychotic mode. We are then confronted with the stumbling-blocks of associativity, which have to do with defects in symbolization, splitting and ruptures of thought. And yet the analyst works here as elsewhere with his own associations and tracks the associativity that sometimes occurs in the patient’s discourse, revealing elaborative capacities on which he can rely.