Strategies for syntactic processing
The preceding sections have presented summaries of a few issues in sentence complexity and in syntactic marking that have implications for the syntactic processing of sentences. How, then, do we carry out the processing of sentences How do we build sentences using the input words It seems that there are some basic observations that need to be taken into account.
One is that discontinuous constituents are difficult to process. Compare the sentences in (10.32)– (10.34_. The more material that intervenes between a and, the more difficult it is to process the sentence.
Of course, sentences like (10.34), though grammatically permissible, would normally be avoided or expressed differently. The problem with such sentences is that when we encounter the word it is disconnected from what we have read or heard just prior to that point. It seems that we package up the constituents that we read or hear, and a no longer belongs together as a single constituent.
Another observation is that native speakers seem to have clear preferences in the structures they assign to sentences. The same is probably true of non-native speakers, though less is known about their approaches to the processing of sentences (Frenck-Mestre, 2005 ; Nitschke, Kidd & Serratrice, 2010). Although (10.35) is unambiguous, it is problematic. Most readers do a double-take, and have to go back and read the sentence again to get the right analysis. By contrast, (10.36) is ambiguous, but it is often hard to detect the ambiguity which involves the relationship of o s to the preceding material – did Pat buy the book for Chris or had I been trying to find it for Chris?)
It is argued that these effects come about because human sentence processing automatically builds the words of a sentence into particular preferred structures, so that for instance the adverb ese a at the end of (10.35) is attached to the most recent verb phrase, despite the fact that the phrase is marked for future but ese a indicates past.
