| I examine the role of context in understanding noun-noun combinations. The comprehension of noun-noun combinations depends on the identification of the implicit relation, or relationship, that binds the two concepts. Research on context effects has typically employed experimenter-written passages that provide both a referent for the conceptual combination and an explicit relationship between its constituents. In addition, the target combinations in these typical paradigms are semantically indeterminate, or have multiple potential interpretations. Two text analyses evaluated the validity of these standard assumptions by comparing experimental paradigms to actual discourse from newspaper and magazine articles. The first analysis investigated whether the constituents of the phrase preceded the phrase in context; the second established whether a referent or relation preceded the phrase. The analyses revealed that in actual discourse, context usually does not provide the phrase's constituents, nor a referent or relation for the phrase. Two experiments compared comprehension of discourse phrases and experimental phrases both within their respective contexts and in isolation. Experimental phrases were better understood in context than in isolation, while discourse phrases did not benefit significantly from context. |
