Psycholinguistic accounts of language acquisition coincide on the major role that attention plays in converting input to intake (Gass, 1997). According to Long (1991) and Swain (1998), attention to form helps learners notice the gaps that exist in their interlanguage grammar and can facilitate intake derivation. However, attention to form should be accompanied by attention to meaning, and form–meaning connections require great mental effort. Because the cognitive resources needed to respond to this demand are limited (Just & Carpenter, 1992), learners have to select which aspects of the input they will process (Gass et al., 2003).

VanPatten (1996, 2004b) has shown that learners tend to process features of the language that carry more meaning (e.g. content words) before those that are less meaningful (e.g. grammatical forms) (see also Clahsen & Felser, 2006). VanPatten (2004b) also notes that, when processing meaningful forms, learners prefer to process non-redundant over redundant elements (see also Lee, 1999, 2002; Lee et al., 1997; Leeser, 2004; Musumeci, 1989; Rossomondo, 2003). Han and Peverly (2007) (see also Park & Han, this volume) expand VanPatten's input processing theory, arguing that, whereas learners with previous L2 knowledge prefer to process non-redundant forms first, those with no existing L2 knowledge attend to grammatical forms independently of their communicative redundancy.

Despite the substantial body of literature devoted to investigating whether L2 learners are able to attend to redundant grammatical forms, most studies have examined this question employing off-line techniques and, thus, can only provide circumstantial evidence about L2 processing. In order to explore what happens during real-time processing, studies need to adopt on-line techniques that measure attention as processing unfolds. In 2001, Sagarra and Dussias began to cover this lacuna by using eye movement records to examine attention to redundant grammatical forms in intermediate L2 learners.