Age of acquisition (AoA), also called age of onset (AoO), is the age at which a learner begins to receive substantial input in a second language. It is one of the most studied and most misunderstood variables in SLA: easy to measure, strongly correlated with outcomes in naturalistic settings, far less straightforward in instructed foreign-language classrooms.
In immigrant and long-residence naturalistic contexts, AoA shows a strong and consistent negative correlation with nativelike ultimate attainment. Later starters reach lower plateaus, with the clearest break around puberty on phonology and a more gradual decline on morphosyntax. This pattern is the empirical ground for the Critical Period Hypothesis and for Bley-Vroman's Fundamental Difference Hypothesis.
The Stockholm programme of Kenneth Hyltenstam, Niclas Abrahamsson, and Bylund has sharpened the finding by distinguishing near-native from nativelike. When late starters who pass listener judgement are scrutinised on sensitive grammatical and phonetic tasks, virtually none reach the native range, even at ages of onset in late childhood.
Inside the classroom, the naturalistic story does not transfer cleanly. Carmen Muñoz's Barcelona Age Factor (BAF) Project, controlling for hours of exposure, found that older starters outperform younger starters on rate of acquisition at equivalent input. "Earlier is better" holds for naturalistic immersion, not for a few hours a week of foreign-language classes in a school system. The policy move to push English instruction into earlier grades is, on Muñoz's evidence, poorly justified.
Two common misreadings need resisting:
AoA sounds simple but is harder than it looks. Decisions about what counts as "beginning to acquire" — first exposure? first sustained input? arrival in the host country? — produce different numbers for the same learner. Studies that find different AoA effects sometimes disagree because they have operationalised the variable differently.