摘要: This paper discusses ways of determining whether the human parser is serial maintaining at most, one structural interpretation each parse state, or it parallel, more than in least some circumstances. We make four points. The first two counterclaims made by Lewis (2000): (1) that availability alternative structures should not vary as a function disambiguating material ranked parallel models; and (2) models predict slow down during ambiguous region for syntactically structures. Our other points concern potential methods seeking experimental evidence relevant to serial/parallel question. discuss effects plausibility secondary structure (Pearlmutter & Mendelsohn, 1999) suggest examining distribution reaction times region.