Abstract: Three experiments were conducted to investigate relationships between time metaphor and perceptual representation system. In each experiment, time classifiers, which were displayed in pairs on the left and right of the screen respectively, were used as targets. All target pairs were represented either the same or different length of time period (e.g. “year” and “year”, or “year” and “second”). Therefore, there were same-period targets and different-period targets. In each different-period target pair, the one represented longer time period was called longer-time classifiers, whereas the other one was shorter-time classier. In Experiment 1 and Experiment 2, sizes of the targets were horizontally (Experiment 1) or vertical (Experiment 2) prolonged in two different levels (200% and 400%). In Experiment 3, targets were displayed with gray background, including double length and triple length conditions. It is called metaphor-consistent condition, if longer-time classifiers in this condition were physically longer, taller, or with longer background than the other one; by contrast, it is called metaphor-inconsistent condition, if longer-time classifiers were physically shorter, or with shorter background. Participants were instructed to judge which time classifier represented longer time period. In experiment 1 (N=30), semantic processing was inhibited in metaphor-inconsistent condition rather than metaphor-consistent condition with same-period target pairs, and an expected facilitation occurred in metaphor-consistent condition when time classifiers were physically different. In addition, facilitation for the condition with both metaphor consistency and length consistency is much larger than the rest of the conditions. It might be caused by one of the following plausible reasons: (1) perceptual representation system did affect the process of time metaphor; (2) participant used a pattern strategy that the same-period targets were much more easily judged than different-period targets; (3) or there was response superiority for such a condition that it was responded with “SPACE” whereas the other conditions were responded with either “F” or “J”. In Experiment 2 (N=30), one target was or was not physically taller than the other one in each pair. Results showed no inhibition from metaphor-inconsistent time classifiers. It indicated that time metaphor was affected by usage of language, with which long and short, but not tall and short, are used to describe the time period in Chinese. Such results can be also use to deny the second and the third plausible reasons rose in Experiment 1. In Experiment 3 (N=30), difficulties of the task were decreased by controlling the length factor with changing the background but not changing the physical length of words. Facilitation of the different-period targets was again found in metaphor-consistent condition. Inhibition and facilitation from the perceptual representation might not be activated at the same time.
In short, it is claimed that the information of length can be activated when processing time classifiers. Time metaphor and perceptual representation might be saved together in long term memory.

Key words: metaphor, perceptual representation theory, time classifier