This is an original preprint of an article published in The International Journal of Psychoanalysis on 17 Apr 2020, available online: hps://www.tandfonline.com/doi/full/10.1080/00207578.2020.1726713 . The supplement of this version has not been published yet. The bringing together of elements that have apparently no connecon in fact or in logic in such a way that their connecon is displayed and an unsuspected coherence revealed…is characterisc of Ps↔D [the paranoid-schizoid and depressive posions interplay] (Bion 1989) On the Logic of the Unconscious/ Amit Saad Abstract Freud argues that the laws of classical logic do not hold true with regard to the unconscious. The aim of this paper is to provide an alternave logic that is consistent with Freud’s concepon of the unconscious. This aim is achieved in two steps. First, I argue that most Freudian features of the unconscious are consequences of the fact that unconscious thoughts in the Freudian sense are negaonless. Then I suggest a formal logical system that lacks the negaon symbol, and is known as negaonless-logic, with which to formulate the logic of the unconscious. I demonstrate that this suggeson provides a beer understanding of various psychological phenomena, such as projecon, reacon-formaon, delusion, and the misconstrual of the empty set in childhood. Finally, I argue that negaonless-logic provides soluons to certain shortcomings in Mae-Blanco’s formulaon of the logic of the unconscious. 1. Introducon Psychological accounts typically include proposions such as John thinks that apples are tasty, or Jane wished that she would get the job. The general form of such proposions is: (*) A thinks/believes/wishes/hopes etc. that p. where A is a subject and p is a proposion. Usually we jusfy such claims based on observaons of the subject’s behaviour and uerances (Denne 1989; Hacker 1990; Wigenstein 1953). For example, if John prefers apples over bananas, it would usually be jusfied to assert that John thinks that apples are taser than bananas. Before Freud, it was assumed that a claim like (*) enjoys first-person authority, i.e. 1