Skip to main content
. 2019 Aug 21;6(8):191023. doi: 10.1098/rsos.191023

Figure 1.

Figure 1.

Sketch of the database and analysis. (a) We show the waveform of a speech sample and the alignment for three linguistic levels of symbolic transcription: phonemes, words and BG. (b) We showcase how the same symbolic unit (the word okay) may show a wide diversity within speech communication (number of phonemes, phoneme type, duration, etc.). We combine all this information in order to (c) characterize statistical patterns and linguistic laws in both symbolic and physical magnitudes at three different levels, and discuss the relationship between them.