Pérez, M. &. & Sáez, L.
University of Murcia, Murcia, Spain
Lexicon is governed by the order at which words are learnt so that words that are learned early are accessed and retrieved faster than words acquired later in life. This is called the age of acquisition (AoA) or, as it is preferred nowadays, the order of acquisition (OoA) effect and it has been observed and well established across a wide range of tasks and population samples.
The main aim of the present work is to assess whether or not an OoA effect appears when people learn a limited set of words in a natural setting. The manipulation consisted in including a number of words in different moments of a German course for Spanish students. Twenty words were trained from the beginning of the course (henceforth, early words). Four weeks later a new set of words were trained (henceforth, late words), and the early words were also rehearsed. Word length, bigram frequency, neighbourhood density (in Spanish) and lexical frequency in German were matched among sets. All words were object names. At the end of training, each early word was trained 17 across 13 sessions, and each late word was trained 17 times across 11 sessions. Training exercises covered phonology, orthography and semantics of words (e.g., filling the gaps, word spelling, and spoken word - picture association)
OoA effects were tested by using picture naming and visual lexical decision tasks at 7 days after the last training session. A main OoA effect was found in each task. However, the magnitude of the effect varied across tasks. The results extend OoA effects found in lab tasks to a more natural L2 acquisition context. Moreover, since meanings of trained words were previously known by participants, the results support mainly mapping hypothesis, which claims that OoA effects reside in the connections between word-forms and meanings.