Text problems
Several examples illustrate decoding errors arising possibly due to deficient vocabulary
knowledge. In particular, idiomatic language, proper nouns, and low frequency words were
problematic. What the examples reveal is that a known word was matched instead of the
actual word in the speech stream, with most, if not all, of the choices learners adhered to
being semantically inappropriate. Field (2004) also found this propensity among learners to
match, sometimes very approximately, a problem word with a known one, and suggested
that such decoding errors occur due to a known (and more frequent) word possibly distorting
‘bottomup’ perceptual evidence and overriding ‘topdown’ contextual (and cotextual)
information, as well as because such counterindications were potentially being ignored
(e.g., the cotext ‘by a tractor’ in Example 3) or not attended to (e.g., visual information). Additionally, the partial responses given for some clauses in this study indicate that cotextual
information was not always decoded and available to inform decisionmaking
regarding word appropriacy, and therefore it may be that a known word could not be
rejected in such cases, regardless of it being inappropriate. Furthermore, it is evident that
learners in the given examples had adhered to an incorrect match across the two
opportunities to listen. While this may be due to an inability to decode the correct word
across the two listenings, Field (2008c) offers an alternative interpretation in terms of the
“perseveration effect” (p. 48). That is, once a candidate had been chosen as a good match in
the first listening, learners were reluctant to engage in the competition process once more, in
spite of encountering evidence in the second listening contradicting their initial match. This
may be due to learners not having enough attentional capacity to reconsider competitors
(Field, 2008c).
Intrusion problems
Among the examples, it was apparent that the learners’ L1 influenced the decoding process. Firstly, some learners encountered minimal pair interference in terms of being unable to
discriminate between /b/ and /v/, preferring to assimilate both sounds to Japanese /b/. This
difficulty is not uncommon for Japanese learners of English, though it becomes less evident
with more experience of the L2 (Brown, 2000). Nonetheless, it seemed to be relatively
persistent for the given advancedlevel learners in this study. In addition, a welldocumented