Fig 8. Non-Markovian repeated syllables are loudest and evoke the largest HVC auditory responses.

a: Mean ±s.d. amplitude waveforms for a non-Markovian repeated syllable (black), a Markov repeated syllable (red), and an intro note (grey) from the songs of one bird. b: Mean ±s.e. normalized peak amplitudes of song vocal elements. Intro notes (Intro), non-repeated syllables (NR), Markov-repeated syllables (MR, peak repeat number = 1), and non-Markovian repeated syllables (nMR, peak repeat number > 1). non-Markovian repeated syllables are significantly louder than other vocalizations (p < 10⁻³, p < 0.01, Wilcoxon sign-rank test, Bonferroni corrected for m = 3 comparisons). c: Scatter plot of normalized auditory responses to a syllable as a function of the normalized amplitude of that syllable. Black line is from regression (R ² = 0.30;p < 10⁻³, N = 40 syllables). d: Paired comparison of normalized auditory responses to non-repeated syllables (NR) and non-Markovian repeated syllables (nMR). Repeated syllables illicit larger auditory responses. (p < 0.01, Wilcoxon sign-rank test, N = 11 sites). Circles: data; square: median.