Exercise 10.4

In the table above, the four regions of the sentences have been labelled pre-amb for the region before the ambiguous phrase of interest, amb for the ambiguous phrase itself, then post-amb for the region immediately after the ambiguous phrase and rem for the remainder. Notice that the two sentences are identical up to and including the amb region. In the self-paced reading study, there was no difference in reading times for the two sentences over the first two regions. For the post-amb region, however (and to a lesser extent for the rem region), reading times were slower for the sentence in 1 than for the sentence in 2, even when word-length differences are accounted for. The claim is that this is because readers prefer to interpret the category ambiguous word like as a preposition, because that is a more frequently-encountered usage than its verb use. In 2, the post-amb region makes it clear that like is here a verb, and this results in a 'garden path' experience and delays in processing because revision is required.

Note that the result above is for native speakers. The same study found that learners of English did not show the same preference for like as a preposition rather than as a verb. This reflects both the patterns of usage in corpora of non-native writing, and the way in which like is typically taught to second-language learners, i.e. as a verb and rarely explicitly as a preposition.

For more on this, see:
Vine, E. W., & Warren, P. (2012). Corpus, coursebook and psycholinguistic evidence on use and concept: The case of category ambiguity. In S. Hoffman, P. Rayson & G. Leech (Eds.), English Corpus Linguistics: Looking Back, Moving Forward (pp. 239-247). Amsterdam: Rodopi.