Identifying and interpreting blends: An experimental approach

Factors affecting the identification & evaluation of novel blends are investigated experimentally by presenting students in an undergraduate linguistics course (N unspecified) with two questionnaires. Questionnaire 1 asks Ss to identify the contributing words to a set of blends, specify meanings if not obvious, & rate each blend as a good/bad word on a 1-5 scale. Questionnaire 2 tests the blends in their natural context & varies rating instructions in terms of comparisons. Ss also had to indicate all blends they had previously encountered. Results of statistical analyses tend to confirm hypotheses that (1) blends are more easily identified in context; (2) the more material from the target word is present, the easier the blend is to identify; (3) the higher the frequency of the target word, the easier it is to identify; (4) the fewer the number of words in the neighborhood of the target word, the easier it is to identify; & (5) if one part of a blend is identified, its semantics will be relevant to the identification of the other part. An ordered sequence of strategies used to segment & interpret blends is proposed. 8 Tables, 1 Appendix, 42 References. Adapted from the source document