Referential Coding Contributes to the Horizontal SMARC Effect Yang Seok Cho Korea University Gi Yeul Bae Johns Hopkins University Robert W. Proctor Purdue University The present study tested whether coding of tone pitch relative to a referent contributes to the correspon- dence effect between the pitch height of an auditory stimulus and the location of a lateralized response. When left–right responses are mapped to high or low pitch tones, performance is better with the high–right/low–left mapping than with the opposite mapping, a phenomenon called the horizontal SMARC effect. However, when pitch height is task irrelevant, the horizontal SMARC effect occurs only for musicians. In Experiment 1, nonmusicians performed a pitch discrimination task, and the SMARC effect was evident regardless of whether a referent tone was presented. However, in Experiment 2, for a timbre-judgment task, nonmusicians showed a SMARC effect only when a referent tone was presented, whereas musicians showed a SMARC effect that did not interact with presence/absence of the referent. Dependence of the SMARC effect for nonmusicians on a reference tone was replicated in Experiment 3, in which judgments of the color of a visual stimulus were made in the presence of a concurrent high- or low-pitched pure tone. These results suggest that referential coding of pitch height is a key determinant for the horizontal SMARC effect when pitch height is irrelevant to the task. Keywords: auditory compatibility, polarity coding, Simon effect, SMARC effect, SPARC effect Spatial correspondence between stimulus and response alterna- tives has a strong influence on the response-selection process (e.g., Proctor & Vu, 2006). For example, left–right responses are faster and more accurate when the stimulus and response correspond spatially than when they do not, regardless of whether the stimulus location is task relevant or irrelevant (see Hommel & Prinz, 1997). This stimulus–response compatibility (SRC) effect is also obtained with various other types of stimulus and response sets. For in- stance, when people make an up or down response to a high or low pitched tone, performance is usually better for the mapping of the high tone to the up response and the low tone to the down response than for the opposite mapping. This phenomenon is called the spatial musical association of response code (SMARC) effect (Rusconi, Kwan, Giordano, Umilta `, & Butterworth, 2006) or, alternatively, the spatial pitch association of response code (SPARC) effect (Lidji, Kolinsky, Lochy, & Morais, 2007). This SMARC effect also occurs when responses are made to the timbre of the tone (Lidji et al., 2007; Rusconi et al., 2006) or the color of a visual stimulus (Nishimura & Yokosawa, 2009), rather than to the tone’s pitch, much like the spatial SRC effect obtained when stimulus location is task irrelevant (called the Simon effect; e.g., Simon & Rudell, 1967). This vertical SMARC effect is thought to be due to associations of pitch height with space, much like the more widely studied spatial–numerical association of response codes (SNARC) effect is attributed to associations of number magnitude (small or large) with space (left or right; Lidji et al., 2007; Rusconi et al., 2006). That is, pitch height is spatially represented as a vertically ordered line with low pitch tones located on the lower part and high pitch tones on the upper part of the mental pitch line. Rusconi et al. (2006) proposed that the “human cognitive system maps pitch onto a mental representation of space” (p. 126). Because high pitched tones are located on the upper part and low pitched tones on the lower part of the vertically arrayed mental representation of pitch height, spatial correspondence between the musical pitch and upper–lower responses occurs. Lidji et al. (2007) suggested that the compatibility effect between pitch height and vertical response is obtained because of an intrinsic characteristic of the tone pitch. The SMARC effect is also obtained with a horizontal response set. When Lidji et al. (2007) had participants perform a pitch comparison task with left–right responses in their Experiment 2, performance was better with the mapping of “high–right/low–left” than with that of “high–left/low–right” for both musicians and nonmusicians. Rusconi et al. (2006) also found a 16.5-ms “high– right/low–left” advantage in their Experiment 1, although it was not statistically significant. They suggested that the SMARC effect obtained with the lateralized response set is consistent with the view that the mental representation of pitch is spatially multidi- mensional (Mudd, 1963), having a horizontal representation as well as a vertical one. In addition, Rusconi et al. and Nishimura This article was published Online First December 5, 2011. Yang Seok Cho, Department of Psychology, Korea University, Seoul, Korea; Gi Yeul Bae, Department of Psychological and Brain Sciences, Johns Hopkins University; Robert W. Proctor, Department of Psycholog- ical Sciences, Purdue University. This research was supported in part by the Korea Science and Engineer- ing Foundation (KOSEF) Grant, funded by the Korean Government (MOST-2006-05110), and a Korea University Grant. Correspondence concerning this article should be addressed to Yang Seok Cho, Department of Psychology, Anam-dong Seongbuk-Gu, Seoul, Korea 136-701. E-mail: yscho_psych@korea.ac.kr Journal of Experimental Psychology: © 2011 American Psychological Association Human Perception and Performance 2012, Vol. 38, No. 3, 726 –734 0096-1523/12/$12.00 DOI: 10.1037/a0026157 726