Discovery Misattribution: When Solving Is Confused With Remembering Sonya Dougal New York University Jonathan W. Schooler University of British Columbia This study explored the discovery misattribution hypothesis, which posits that the experience of solving an insight problem can be confused with recognition. In Experiment 1, solutions to successfully solved anagrams were more likely to be judged as old on a recognition test than were solutions to unsolved anagrams regardless of whether they had been studied. Experiment 2 demonstrated that anagram solving can increase the proportion of “old” judgments relative to words presented outright. Experiment 3 revealed that under certain conditions, solving anagrams influences the proportion of “old” judgments to unrelated items immediately following the solved item. In Experiment 4, the effect of solving was reduced by the introduction of a delay between solving the anagrams and the recognition judgments. Finally, Experiments 5 and 6 demonstrated that anagram solving leads to an illusion of recollection. Keywords: recognition, memory misattribution, recollection, insight problem There is something very similar about the experience of recol- lection and that of discovery. In both cases, a thought comes to mind with a compelling sense of truth. In the case of recollection, the truth value of a remembered thought results from its associa- tion to the past, whereas in the case of discovery, veracity stems from the compelling manner in which a generated idea solves a problem. The subjective parallels between recollection and discov- ery raise the possibility that the two experiences could be confused with one another. Several studies have investigated instances in which remembering is mistaken for discovery, leading to cryptom- nesia, or unconscious plagiarism (e.g., Brown & Halliday, 1991; Marsh, Landau, & Hicks, 1997). However, far less attention has been given to the possibility that discovering a solution may lead to false recollection. In this article, we present data supporting the hypothesis that the experience of successfully solving a problem can be confused with the experience of recollection, thus leading to a “discovery misattribution.” Considerable research has documented the ways in which mis- attribution processes contribute to false memories (e.g., Jacoby, Kelley, & Dywan, 1989; Johnson, Hashtroudi, & Lindsay, 1993). One source of evidence for misattribution processes in memory comes from situations in which perceptual fluency is confused with item familiarity. For example, Jacoby and Whitehouse (1989) found that increasing the perceptual fluency of items on a recog- nition test by preceding them with subliminally presented primes increased the proportion of “old” judgments to test items. In contrast, Whittlesea, Jacoby, and Girard (1990) found that decreas- ing the perceptual fluency of items on a recognition test by degrading their perceptual clarity reduced the proportion of “old” judgments to test items. In the above cases, the perceptual fluency of the recognition items influenced the experience of familiarity, which led to memory misattributions. Of note, these studies dem- onstrate that nonmemorial phenomenological states may influence memory judgments when those states resemble an experience associated with remembering. In the case of discovery misattributions, the question arises as to whether there are phenomenological states associated with prob- lem solving that may be confused with the experience of recollec- tion. One likely candidate is the well-documented sense of “aha” that is associated with discovering the solution to insight problems (e.g., Bowden, Jung-Beeman, Fleck, & Kounios, 2005). The “aha” experience represents an emotional state (Gick & Lockhart, 1995) associated with the surprise of suddenly discovering the solution to an insight problem such as solving an anagram (Metcalfe, 1986). Indeed, Jung-Beeman et al. (2004) found that solving problems that elicit the “aha” experience is associated with activation of the amygdala, a brain region previously shown to be involved in emotional arousal (e.g., Phelps, 2004). We hypothesize that the emotional arousal elicited by successfully solving insight problems may be used to inform subsequent memory judgments (e.g., Schachter & Singer, 1962). Although no study has specifically examined whether the “aha” experience associated with solving problems might be confused with recollection, a number of studies have demonstrated that the related experience of surprise influences recognition judgments. For example, Whittlesea and Williams (1998) demonstrated that on a recognition test including natural words, pseudohomophones of the words (e.g., Frog spelled Phrawg), and nonwords, the pseudohomophones were more likely to be classified as old than both natural words and nonwords. However, reading times for the natural words were faster than for the pseudohomophones, indi- Sonya Dougal, Department of Psychology, New York University; Jonathan W. Schooler, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada. The authors contributed equally to this research. Portions of this re- search were presented at the Annual Meeting of the Psychonomic Society, Dallas, Texas, November 1998, and the Third International Conference on Memory, Valencia, Spain, July 2001. This research was supported by a Mellon Fellowship to Sonya Dougal. We are indebted to Marcia Johnson, whose comments motivated the design and analysis of Experiment 1, and Martin Conway, who suggested the use of a delay manipulation in Experiment 4. Correspondence concerning this article should be addressed to Sonya Dougal, Department of Psychology, New York University, 6 Washington Place, Room 865, New York, NY 10003. E-mail: sonya.dougal@nyu.edu Journal of Experimental Psychology: General Copyright 2007 by the American Psychological Association 2007, Vol. 136, No. 4, 577–592 0096-3445/07/$12.00 DOI: 10.1037/0096-3445.136.4.577 577