Figure 1. Explained variance ( R 2 ) for each semantic scale with the stepwise multiple linear regression (in solid blue) and with the distance metric learning on neuromimetic representations (in dashed red). More

Figure 1. Example of musical duo (No. 4) performing during the experiment. More

Figure 2. Marker and joint locations. a) Anterior and posterior views of the marker placement on the participants’ bodies; b) Anterior view of the marker locations as stick figure illustration; c) Anterior view of the locations of the secondary markers/joints used in the analysis. More

Figure 3. Movement features of duo musicians playing vs. acting the solo and accompaniment part. A. Quantity of Motion (head and shoulders). B. Fluidity (head and shoulders). C. Bounding Rectangle (head and shoulders). D. Shoulder Speed along x-, y-, and z-axes. More

Figure 4. Observers’ average dwell time on duo musicians performing congruently (musical and behavioral roles aligned) and incongruently (musical and behavioral roles opposed). Note : Behav. = Behavioral. Acc. = Accompaniment. More

Figure 1. A sample of pitch contour + timing patterns, like those used by Creel (2012) . While W and X share rhythmic patterns, Creel (2012) treated them as different motifs because their contours differed. Neither Creel (2012) nor the current study treat Y and Z as the same, even though they... More

Figure 2. Experiment 1, two example melodies (top lines) with a 3/4 context (lines 2-3) and a 6/8 context (lines 4-5). Only one context (that is, accompaniment) was heard at a time. More

Figure 3. Experiment 1, metrical restoration effect as a function of timbre and rate variability. (a) Raw goodness-of fit ratings; (b) metrical restoration score. The first four bars in (a) correspond to the first bar in (b). Throughout, error bars are standard errors. ** p < .01, * p < ... More

Figure 4. Experiment 2, metrical restoration (dark gray), with overall effects from Experiment 1 for comparison (light gray). * p < .05, ** p < .01, *** p < .001 More

Figure 5. Experiment 3 metrical restoration (dark gray), with Experiment 2’s analogous conditions (light gray). *** p < .001, ** p < .01 More

Figure 6. Experiment 4, metrical restoration. ** p < .01, * p < .05 More

Figure 7. Experiment 5, metrical restoration effect. *** p < .001, ** p < .01 More

Figure 8. Experiment 6, metrical restoration. * p < .05 More

Figure 9. Summary of effects in Experiments 2-6, showing how rhythmic patterns were distributed across melodies. For each experiment, only “conflict” conditions are depicted—those where melody pairs with similar or shared rhythmic patterns, such as Melodies 1 and 2, mismatched in the meters they o... More

Figure B1. Control experiment, metrical probe ratings, ± standard errors. More

Figure B2. Experiment 2, probe ratings for the most 6/8-biased and the most 3/4-biased melody according to the control study. Experiment 2 ratings for the 6/8-biased melody were overall higher for 6/8, and the reverse was true for the 3/4 biased melody. Nonetheless, exposure effects were evident f... More

Figure C1. Experiment 1, six example melodies. Measure-long rhythmic patterns are notated numerically above each measure. More

Figure C2. Experiment 2, first half of each of the eight melodies. Rhythmic patterns are marked numerically over each measure. For present purposes, dotted half notes were not considered to be shared patterns. More

Figure C3. Experiment 3, example pair of melodies with shared rhythmic patterns (4 1 1, 3 1 1 1), first 8 measures of each melody. In the matching condition, both would be heard with 3/4 accompaniments (or both 6/8). In the conflicting condition, the first would be heard in 3/4 and the second ... More

Figure C4. Experiment 4, example melodies with changed rhythmic patterns (first 8 measures of each melody). More