Can adults learn absolute pitch? Evaluating the “Melody Triggers” method

Despite over a century of research into absolute pitch (AP), musicians and scientists alike still cannot agree on what it is or how it develops. Some researchers maintain that it is innate, although there is little compelling evidence to support this. Many studies suggest that it can be fostered through early musical training.

Deutsch (2013) proposes that AP learning is subject to the same critical period as phonological development and is supported by the same neural mechanisms. However, evidence suggests that maturational constraints on phonology acquisition are considerably less rigid than those on syntax (Curtiss, 1977; Fattal et al., 2011; Friedmann & Rusou, 2015), and that in the presence of a native language adults can attain nativelike phonology under special conditions (Bongaerts, 1999; cf. Reetzke et al., 2018). As such, AP acquisition should likewise remain possible under this hypothesis. Yet a clear disparity remains in ultimate AP attainment between adults and children. This thesis therefore proposes an alternative account: the AP Fundamental Difference Hypothesis. According to this view, AP develops primarily through procedural memory in childhood, whereas adult acquisition relies more heavily on explicit memory. Although adults can proceduralise pitch associations that initially require conscious retrieval, what becomes proceduralised is typically less complete and is consolidated through less reliable processes, reflecting learner-specific strategies and goals. Importantly, this suggests that although adult learners can achieve very high levels of proficiency, attaining truly "native-like" AP is—if possible at all—extremely difficult and exceptionally rare. This perspective accounts for possessors who exhibit only partial AP, usually restricted to particular timbres or pitch ranges and often aided by motor memory (Ross, 2004; Ross et al., 2005). Rather than reflecting a fundamentally different capacity from "true AP," such cases represent lower proficiency across certain dimensions of the same ability. This interpretation is supported by findings suggesting a continuous distribution of AP performance in the population rather than discrete categories (Bermudez & Zatorre, 2009; Miyazaki et al., 2018).

Researchers have observed differences between AP possessors and non-possessors in brain anatomy, brain function, neural oscillations, and the timing of brain activity. However, it is unsurprising that these groups process pitch differently, and the anatomical differences are likely attributable to learning-induced neuroplasticity (Zatorre et al., 2012; Meyer et al., 2012). Three recent studies provide suggestive evidence that adults can learn AP (Bittrich et al., 2015; Van Hedger et al., 2019; Wong et al., 2020a). Moreover, findings that AP performance is influenced by recent experience suggest that its degree and quality are not rigidly fixed—even for "native" possessors (Bahr et al., 2005). Additionally, there is evidence that even adults without AP can demonstrate "implicit AP" through their ability to accurately sing the opening pitches of well-known songs (Levitin, 1994).

In light of this, the present study tests the hypothesis that adults can learn AP by training to associate pitch labels with "Melody Triggers" (Aruffo, 2007b). It also investigates two further hypotheses: (a) that identifying pitches within a musical listening task poses a different—and greater—challenge than identifying isolated pitches; and (b) that this dimension of AP can be trained. Finally, the study proposes a new multidimensional model of AP, informed by models from Bahr et al. (2005) and Takeuchi & Hulse (1993).

Data from an online survey and two experiments provide strong support for the study's hypotheses. The results endorse the efficacy of the "Melody Triggers" method and show that identifying pitches within musical contexts presents a greater—or at least different—challenge than standard pitch identification testing, yet one that can be overcome with training. These findings underscore the need for a more ecologically valid AP test and raise important questions about how existing findings based on standard pitch identification tests should be interpreted. The survey data also reveal that many musicians who exhibit AP limited to timbre—developed through extensive practice on a specific instrument—interpret their ability as a form of relative pitch rather than AP, suggesting that more musicians may possess a substantial degree of AP than is typically acknowledged.

© 2025 Sam Leak. All rights reserved.

This thesis was awarded by the University of Cambridge and is available via DOI link.