In the second experiment it was hypothesized that the initial f0 peak is only sensitive to length manipulations of the first constituent.
The expert literary readers were found to assign significantly higher ratings to all versions of the manipulated poems than the novice readers.