Implicit voice learning through discrimination outperforms explicit listen-and-memorize tasks

Voice learning primarily occurs implicitly in everyday situations—as an incidental by-product of other activities such as participating in conversation or listening to voices in the media. Most research on voice learning is conducted in laboratory settings, where participants are explicitly instructed to attend to and memorize voices for later recognition. Yet, the impact of task awareness (awareness regarding the goal of voice learning) on voice recognition performance remains poorly understood. To address this gap, we conducted a study comparing two voice-learning modalities: explicit learning, instructing participants to listen to and memorize voices for later recognition, and implicit learning, based on exposure during a voice discrimination task, without awareness of a subsequent recognition test. After both exposure phases, participants completed a surreptitious old–new voice recognition task. To further examine whether task awareness is modulated by voice load (number of voice identities introduced in the experiment), we implemented both a simple and a challenging version of the experiment. We found that, irrespective of voice load, implicit learning through participation in a discrimination task, resulted in higher recognition performance than explicit listen-and-memorize training. These findings suggest that highly demanding explicit listen-and-memorize tasks may benefit from incorporating ecologically valid familiarization paradigms, such as voice discrimination. We discuss the implications of our findings in relation to previous empirical research and their relevance for forensic applications.