A revised physical-correlate theory relating softness of a tone to perceived distance

Abstract

The purpose of this study was to obtain perceived distance scales of visual and auditory stimuli, and softness scales of auditory stimuli, over various stimulus ranges. Two theories describe the relationship between perceived magnitude and physical intensity. These are: (a) Stevens' psychophysical law, which states that perceived magnitude increases with stimulus intensity raised to a power, with the slope of the function thought to be specific to a given modality due to the activity of the sensory transducers, and (b) Warren's physical-correlate theory (PCT), which accepts the power law but argues that softness judgments are based upon learning a distance effect, and do not derive exclusively from transducer behavior. Previous research on a number of continua indicated that the exponent of the power law varies with stimulus range. An analysis of the assumptions of both theories, and incorporation of the "range effect,'' led to a revised PCT (RPCT). The new theory was shown to relate softness and distance judgments to one another without reliance upon several highly restricting physical laws of the loudness (softness)-distance interaction. The method of magnitude estimation was employed to obtain the several scales. The results were as follows: (1) Perceived distance of a visual target grew as a power function of simulated visual distance. The exponent increased in value with a decrease in stimulus range. In neither range did the value approach that required by the PCT. (2) Softness and perceived distance estimates or a tone in a 14 db and 34 db stimulus range plotted as logarithmic functions of SPL. The stimuli were presented by a loudspeaker and headphones to separate groups of observers. The gain was changed to give the impression of a receding sound source for the distance estimates. The slopes or the logarithmic functions were larger for the 14 db range, in keeping with the predictions of the RPCT. Differences in the slopes between the loudspeaker and headphones conditions were credited to smaller signal-to-noise ratios for the loudspeaker signals. (3) Softness and distance scales for a 60 db range for stimuli presented by headphones approximated the generalized power law. The results accorded with those obtained by Stevens, not with the PCT. The results suggest that the stability of the power law may be due to the use of large stimulus ranges. Scaling procedures relying upon verbal estimates were criticized. Finally, the credibility of the PCT was questioned. Present and previous results indicated that data in support of the PCT for loudness (softness)-distance scaling reflected the biases of a particular experimental procedure. A mathematical model was offered to demonstrate that both logarithmic and power relationships between perceived magnitude and stimulus intensity may be appropriate in magnitude scaling, depending upon stimulus range.