Sentence examples for precursor sentence from inspiring English sources

Experiment 1a showed that the spectral properties of a precursor sentence can influence the categorization of a vowel continuum.

The manipulated vocalic portions were spliced back into the unmanipulated consonantal parts of the original precursor sentence.

The average F1 value over the vocalic portions of the selected precursor sentence was 502 Hz, ranging from roughly 300 Hz (in [i]) to 800 Hz (in [a]).

Experiments 4a and 4b show that reintroducing pitch movement or silent periods in the signal does not change the influence of a precursor sentence on subsequent targets.

The key finding behind the LTAS mechanism is that suppressing or exciting particular frequency regions of a precursor sentence can induce a shift in the categorization of subsequent vowels.

When a precursor sentence had been processed by the same acoustic filter as that used to adjust the spectral tilt of the target stimuli, listeners relied only on the target F2 value.

Watkins (1991) found effects of normalization when listeners categorized targets on an //-to-/ɛ/ continuum that were presented after intelligible precursor sentences that had been filtered.

Watkins and Makin (1994) found that normalization can also be observed with filtered precursor sentences that are played backward and that the amount of normalization is not reduced in this case.

> -wrap-foot> Precursentencesnces were or were not manipulated with respect to several speech characteristics: pitch movement, reversal of syllables, spectral rotation, syllables of equal amplitude, presence versus absence of low amplitude parts (breaks) LTAS measures Figure  5 displays the LTAS plots of the precursors and the endpoint targets, along with their difference LTAS.

The target sounds from each experiment (Experiment 3a for the Experiment 1b stimuli and Experiment 3b for those in Experiment 2) were filtered such that those frequencies that were most pronounced in the precursor sentences would be most suppressed in the target signals.

Another apparent contradiction with our suggested role of familiarity stems from a finding by Remez, Rubin, Nygaard, and Howell (1987), who reported extrinsic normalization effects (also at an ISI of 500 ms) with sine wave replicas of precursor sentences and targets similar to those used by Ladefoged and Broadbent (1957).