Source monitoring in musical tasks : cognitive processes involved in the unconscious plagiarism of musical ideas

Unconscious plagiarism is a serious issue affecting a musician's livelihood, yet little scientific understanding presently informs court decisions on musical plagiarism. In contrast to deliberate copying, unconscious plagiarism occurs when an individual intends to create an original work, but retrieves another person's idea from memory, and mistakes it as their own. The cognitive mechanisms associated with unconscious plagiarism have so far only been tested in verbal creative tasks (e.g., generating category exemplars, or alternate uses for objects). In these studies, unconscious plagiarism was established to be a form of source monitoring error, where externally experienced ideas are confused as internally generated (Johnson, Hashtroudi, & Lindsay, 1993). The likelihood of unconscious plagiarism is greatly increased after improving others' ideas, as this involves similar processes to idea generation, thus, others' ideas become confused as one's own (Stark & Perfect, 2006, 2008; Stark, Perfect, & Newstead, 2005). From this finding it was proposed that the number of high-profile cases of music plagiarism might be explained by the process of reworking and improving a composition (Perfect & Stark, 2008a). The present research investigated this proposal, by replicating Stark and colleagues' studies using a musical composition task. In Studies 1 and 2, we developed and tested a computer-based method for investigating memory for melody, which was then extended to study the cognitive mechanisms associated with unconscious plagiarism in music, using the Brown and Murphy (1989) three-stage paradigm (Studies 3 and 4). This series of studies established two important differences between the cognitive factors which influence plagiarism in musical and verbal tasks. First, in the verbal domain, when recalling or recognising one's own ideas, source monitoring errors are increased specifically by improving others ideas (Perfect & Stark, 2008a, 2012; Stark & Perfect, 2007; Stark et al., 2005). However, in music, we found no evidence that improvement alone increases plagiarism. Instead, re-exposure to ideas, regardless of task, increased plagiarism in music. Second, in verbal studies, the factors which increase plagiarism when generating new ideas are dissociated from those involved in recall and recognition. Generate-new plagiarism is not increased by idea improvement, instead, participants are more likely to plagiarise ideas of positive valence or high quality (Perfect et al., 2009; Perfect & Stark, 2008b; Bink et al, 1999). In music, however, we did not observe this dissociation between generate-new and recognition plagiarism. Instead, re-exposure increased the likelihood of plagiarism in both tasks. Plagiarism of music is therefore more difficult to avoid than verbal content, because no single task can be avoided. In Study 4, we therefore tested an intervention designed to reduce plagiarism. If musical plagiarism depends on exposure, and thus, strength of the idea in memory (Marsh & Bower, 1993), than a distractor task should reduce memory strength for melodies, and thus reduce plagiarism (Glanzer & Cunitz, 1966; Petrusic & Jamieson, 1978). We administered a distractor task consisting of two minutes of randomly generated musical notes at two points during the retention interval of the paradigm. While we observed atypical results in comparison to Study 3, no evidence was found to suggest that the intervention reduced plagiarism. Participant familiarity ratings showed that the intervention did not alter memory strength. The task may have been too brief, or memory for melodies may have been well consolidated after idea generation and elaboration had taken place (Stark & Perfect, 2008; Stark et al., 2005). Across domains, unconscious plagiarism has proven to be difficult to avoid, even when source monitoring is improved (Hollins, Lange, Dennis, & Longmore, 2016; Macrae, Bodenhausen, & Calvini, 1999). The theoretical implications of these studies are discussed in context of the literature on dissociations between implicit and explicit memory. Explicit memory is improved through elaborative processing, but implicit memory is improved through exposure and repetition priming (Schacter & Church, 1992). Musical knowledge is predominantly acquired implicitly, through exposure (Rohrmeier & Rebuschat, 2012). According to the source-monitoring framework (Johnson et al., 1993), where implicit memory is primarily involved, source monitoring processes are not engaged. Thus, our findings in music suggest that this dissociation between implicit and explicit memory systems extends to unconscious plagiarism. Overall, this series of studies identifies that musicians are considerably more vulnerable to unconscious plagiarism than previously understood, and begins to provide a scientific basis for why this is the case. Specifically, our work highlights the limits of source monitoring processes in this domain. Given that unconscious plagiarism in verbal tasks occurs due to errors in source monitoring, reinforcing contextual source cues can protect against plagiarism (Hollins et al., 2016; Macrae et al., 1999). In music, if plagiarism occurs through exposure, no such protection is available. These findings have implications for the legal handling of copyright infringement cases, where no exceptions are presently made for unconscious plagiarism. Policy makers may therefore need to consider the degree to which a musician is held responsible in cases of unconscious plagiarism.