Figure 4

Replication of Ramus et al.²⁴. The horizontal axis represents the percentage of vocalic intervals in a sentence. The vertical axis represents standard deviation of consonantal intervals in a sentence. Five sentences in American English, German, Japanese, and Mandarin Chinese were randomly sampled from 35 (in the case of Mandarin Chinese, 18) sentences used in the intelligibility measure. The analysis results of these languages are represented by coloured symbols. Other languages, i.e., British English, French, and Spanish, are for reference, based on five sentences randomly selected from the same database (represented in black symbols). Error bars represent s.e.m. The red line shows a regression line (ΔC = 0.1862 − 0.0027 × %V; R ² = 0.43), on which the distance among languages can be determined by putting perpendicular lines from the points representing the languages to the scale. The positions of the languages on the scale are ordered according to the rhythm categories with no exception: from left to right, German, American English, and British English can be grouped together as stress-timed languages, French, Mandarin Chinese, and Spanish as syllable-timed, and Japanese as mora-timed.