Is there evidence for unconscious reasoning processes?

Is There Evidence for Unconscious Reasoning Processes? Magda Osman (M.Osman@ucl.ac.uk) Department of Psychology, University College London, Gower Street, London, WC1E 6BT, UK Abstract Current theories of reasoning propose that reasoning is governed by two systems: conscious and unconscious. Conscious processing directs analytical thinking and results in correct responding, while unconscious processing employs heuristics that often leads to poor performance in deductive reasoning tasks. The present study uses a classic propositional task to examine the properties that distinguish conscious from unconscious reasoning. Overall, the study did not find support for dissociable reasoning systems. Instead, the findings suggest that the features exclusively attributed to each system, by dual reasoning theorists, were equally applicable to both. Dual Process Theories At present there are three theories of reasoning that have divided the process into unconscious and conscious components: Evans and Over s (1996) Dual process theory; Slomans s (1966) Two systems theory and Stanovich and West s (2000) Two systems theory. Stanovich and West (1998) present a summary of the general attributes that distinguish conscious from unconscious reasoning processes. They propose that unconscious processes are inaccessible, automatic, inert, non-declarative, and non-verbalizable, while conscious processes are accessible, controllable, declarative, and verbalizable. The different characteristics also imply that the two reasoning systems serve different purposes, result in different responses, and encode information differently. Many studies developed to investigate the different systems originate from Wason s (1966) conditional reasoning task. One reason for this is that the general errors individuals make when solving this task have been the impetus for attributing unconscious mechanisms to reasoning. The aim of the present study is to examine the dual processes theories characterization of deductive reasoning using Wason s (1966) conditional reasoning task. Wason (1966) developed a task (hereafter Wason s selection task) that has now become the mainstay of studies investigating deductive reasoning. It involves a conditional statement if there is a vowel on one side of the card, then there is an even number on the other side . Participants are told that they have to discover whether the statement is true by selecting cards to turn over from an array of four (e.g., E, K, 2, and 5), which are represented in logical notation as (P, ´P, Q, ´Q). The correct solution requires the selection of the E (P) and 5 (´Q) cards, because only this combination provides a means of confirming and falsifying the statement. Typically, only a small proportion of participants solve the task correctly (e.g., 5-10%), while most choose a range of alternative card combinations, the most popular of which is E (P) and 4 (Q). The appeal of this task comes from the robust results it generates, in particular the regularity with which E and 4 cards are selected. The matching bias theory proposed by Evans (1972) and Evans and Lynch (1973) is the most accepted explanation of this phenomenon. They propose that instead of triggering reasoning processes the selection task incites participants to simply match their card choices with those named in the statement. Evans (1972) developed a paradigm to examine this by presenting participants negated versions of the statement. He found that participants still selected P and Q cards irrespective of the presence of negations in the statement, thus leading to the conclusion that the selection task is solved using simple heuristics. A more detailed account proposed by Wason and Evans (1975) explains the underlying processes that motivate matching behavior. They suggested that reasoning is comprised of two dissociated processes, one of which is unconscious and based on quick-fix strategies that are guided by particular preferences for a response (i.e., biases). The second process is conscious and rationalizes behavior, some of which the reasoner has little control over. These proposals were based on findings from protocols studies (Evans & Wason, 1976; Wason & Evans, 1975), which required participants to provide justifications for their card selections. Participants showed a lack of awareness of the actual processes involved in selecting cards, and rationalizations of their behavior were found to be independent of their actual card selections. Following from Wason and Evans protocol studies, a variety of techniques have been developed to uncover unconscious reasoning processes e.g., transfer tasks (e.g., Berry, 1983), and attentional biases (Evans, 1996; Evans, Ball & Brooks, 1987; Roberts, 1997; Roberts & Newton, 2001).