Picked up from seeded input: the Noticing Hypothesis and exposure to binomials

The present study challenges a core tenet of the Noticing Hypothesis, specifically the claim that conscious awareness is always necessary for language learning. While focal awareness can facilitate learning, we demonstrate that lexical sequences, such as novel binomials ( peas or beans ), can be retained in memory through minimal incidental exposure, even when attention is directed elsewhere. To test this, participants completed tasks under high cognitive load, answering questions about everyday topics while being subtly exposed to newly coined binomial phrases embedded in the materials. Crucially, their attention was drawn to content, not form, making deliberate noticing unlikely. In a subsequent surprise recognition task, participants in the experimental groups significantly outperformed control groups, consistently preferring previously encountered variants of binomials. An awareness check confirmed that participants were focused on meaning rather than wording, supporting the interpretation of implicit learning. We take these findings to indicate that initial seeding of formulaic language can occur without conscious attention, providing evidence against the strong version of the Noticing Hypothesis. The study demonstrates the potential of meaning-focused, input-rich contexts to promote the acquisition of formulaic language and suggests that language learning may rely less on deliberate noticing than previously assumed.