Attitudes and approaches toward student error have been a source of debate among second language acquisition (SLA) and second language (L2) writing scholars for more than two decades. The debate has ranged from calls for correction of all student errors to prevent fossilization (e.g., Higgs & Clifford, 1982; Lalande, 1982) to a preference for selective correction that focuses on patterns of error that can be addressed productively (Bates et al., 1993; Ferris, 1995c; Hendrickson, 1978) torecommendations that all error correction be eliminated because it is unnecessary, ineffective and even counterproductive (Cook, 1991; Corder, 1981; Krashen, 1984; Selinker, 1992; Truscott, 1996).

The issue of error treatment (including error analysis, feedback, and instruction) is especially salient in the case of L2 writing classes for two reasons. First, despite disagreement on other points surrounding error correction, there is a fair amount of agreement among researchers on two counts: (1) that accuracy in writing matters to academic and professional audiences and that obvious L2 error may stigmatize writers in some contexts (Ferris&Hedgcock, 1998; Horowitz, 1986; James, 1998; Johns, 1995); and (2) that L2 student writers themselves claim to need and value error feedback from their instructors (Cohen & Cavalcanti, 1990; Ferris, 1995b; Ferris & Roberts, 2001; Hedgcock & Lefkowitz, 1994; Leki, 1991; Truscott, 1996). Second, the research base on the question of whether error feedback helps students to improve in the short run or over time is inadequate as to number of studies and inconsistent as to research design (Ferris, 1999, 2004; Polio, 1997).