تسجيل الدخول

تسجيل

GATING

المؤلف:  John Field

المصدر:  Psycholinguistics

الجزء والصفحة:  P122

2025-08-26

834

GATING

A research method (Grosjean, 1980) which involves presenting subjects with ever-increasing slices of recorded text. A sentence is usually divided into regular time-slices of 20–50 milliseconds, which are termed gates. After hearing the first, subjects report what they think they have heard. They then hear the first gate again followed by the second (i.e. the first 100 milliseconds of the text) and again report what they hear. This continues to the end of the sentence. Gating experiments have demonstrated that on average a word can be recognised within about a third of a second of its onset. In some cases, that is before the word is complete. The finding would appear to support the notion of a uniqueness point as embodied in Cohort Theory: i.e. a point during the utterance of a word where no other item fits the evidence and the word is therefore recognised before it is complete. However, gating has also shown that many short words (especially function words) are not recognised until up to three words after they have been heard. Furthermore, subjects often do not report recognising longer words until some time after their uniqueness points.

Experimenters sometimes ask subjects to record confidence ratings for the words they report. They make a distinction between an isolation point (how much input is needed before a target word is first mentioned), a recognition point, where a confidence level of 80 per cent is reported and (sometimes) a total acceptance point where confidence reaches 100 per cent.

While gating has proved a useful research tool, it involves multiple hearings of parts of the utterance and decision-making which occurs after the stimulus. Some commentators have therefore questioned the extent to which it really taps in to on-line processes.

See also: Uniqueness point

Further reading: Grosjean (1996)