1. Learning Biases Underlie “Universals” in Avian Vocal Sequencing
Logan S. James, Jon T. Sakata 
Current Biology 
Volume 27, Issue 23, Pages e4 (December 2017)
DOI: /j.cub
Copyright © 2017 Elsevier Ltd Terms and Conditions

2. Current Biology 2017 27, 3676-3682.e4DOI: (10.1016/j.cub.2017.10.019)
Copyright © 2017 Elsevier Ltd Terms and Conditions

3. Figure 1
Experimental Tutoring with Randomized and Unbiased Sequences of Syllables
(A) Spectrograms of the five zebra finch “syllables” (i.e., contiguous epochs of sound) in the tutor stimulus set (arbitrarily labeled “a,” “b,” “c,” “d,” and “e”). These five syllables are acoustically distinct from each other and are types commonly observed in zebra finch songs.
(B) Example of tutoring song bout. Five zebra finch syllable exemplars were arranged into 120 possible five-syllable motif variants in which each syllable appears only once. Four distinct motif variants were presented in each song bout. Scale bar represents 250 ms.
(C) A circle plot summarizing transition probabilities between syllables within the motifs of the stimulus set. Outer-ring colors represent the five syllables, and line colors represent transitions from that syllable to others (e.g., blue lines represent transitions from “a” to other syllables). The width of the line is proportional to transition probability. All lines for the stimulus set have the same width because all transitions were heard with equal probabilities.
Current Biology  , e4DOI: ( /j.cub )
Copyright © 2017 Elsevier Ltd Terms and Conditions

4. Figure 2
Skew in the Positioning of Syllables and Positional Variation in Acoustic Features
(A) Representative spectrogram of an experimentally tutored bird’s song, which was composed of the motif “bead.” Scale bar represents 250 ms.
(B) Bar graphs depicting the skew in the likelihood that individual syllables were positioned at the beginning (“Beg”), middle (“Mid”), or end (“End”) of the motif (see STAR Methods). Skew was computed as the standardized difference between the observed number of times a syllable occupied a particular position and the number expected by chance (across 10,000 iterations of the Monte Carlo simulation). Skew is plotted for all five syllables across these three positions, and the significance of the skew was computed using Monte Carlo simulations (p < 0.05). Dashed lines indicate ±1.96 standard deviations (p = 0.05) from the expected (chance) probability (mean of the permutations; solid line). Tutored birds were significantly more likely to position the “b” syllable at the beginning of the motif (p = ), the “e” syllable in the middle (p < ), and the “d” syllable at the end (p = ). Birds were also significantly less likely to produce the “d” syllable at the beginning (p = ) and the “e” syllable at the end of the motif (p = ).
(C–E) The acoustic structure of syllables significantly differed across motif positions for experimentally tutored birds (beginning, middle, or end; n = 200 syllables). Syllables at the beginning or end of the motif were significantly longer in duration (C) and lower in pitch (D; mean frequency) than syllables produced in the middle (p < 0.02 for each). Syllables at the end of the motif were also significantly longer than syllables at the beginning of the motif (p < 0.01). Syllables did not significantly vary in amplitude across positions (E). Amplitude values for each bird are normalized (“norm”; mean subtracted) by the average amplitude of all syllables within his song.
(F–H) The acoustic structure of syllables also significantly varied across motif positions in the songs of birds tutored with species-typical song (n = 267 syllables; see STAR Methods). The duration of syllables significantly varied across positions (F), with syllables at the end of the motif being longer than syllables in the middle or at the beginning of the motif (p < for each). The pitch of syllables varied by position within the motif (G), with syllables at the beginning of the motif being lower in pitch than syllables in the middle or at the end of the motif (p < for each). The mean amplitude of syllables significantly varied across positions in the motif (H), with syllables at the end of the motif being louder than syllables in the middle or at the beginning of the motif and syllables in the middle of the motif being louder than syllables at the beginning of the motif (p < for each). Plotted are mean-subtracted amplitude values (see E).
(C–H) Dots represent median values for each syllable, horizontal lines indicate the means of the distributions, and vertical lines indicate standard errors of the distribution. For all panels, the asterisk (∗) denotes p < See also Figures S1–S4 and Table S1.
Current Biology  , e4DOI: ( /j.cub )
Copyright © 2017 Elsevier Ltd Terms and Conditions

5. Figure 3
Significant Patterning of Syllable Transitions Following Tutoring with Randomized Sequences
(A) Circle plot depicts the syllable transition probabilities within motifs of tutored birds (see Figure 1C for tutoring stimulus).
(B and C) Asymmetries in the directionality of syllable transitions (e.g., difference in transition probabilities from “a” to “e” and from “e” to “a”) within the motifs of tutored birds (B) and tutoring stimulus (C). Line thickness scales with the magnitude of asymmetry in the directionality of syllable transitions (e.g., thick purple line between “e” and “a” in B represents a bias for transitions from “e” to “a” over “a” to “e”).
(D) Overall, there was a significant asymmetry in the directionality of syllable transitions (p = 0.0312; Monte Carlo simulations). Solid line depicts the cumulative distribution of overall asymmetry values found in the simulations. The dashed line represents the 95th percentile from the simulations, and the dot indicates the overall asymmetry value observed across experimentally tutored birds.
Current Biology  , e4DOI: ( /j.cub )
Copyright © 2017 Elsevier Ltd Terms and Conditions

6. Figure 4
Mean Frequency Significantly Alternates across Syllables within Zebra Finch Motifs
(A) A bird’s motif that demonstrates alternation across syllables. Arrows depict the change in mean frequency across adjacent syllables.
(B) The mean frequency of syllables significantly alternated within the motifs of birds tutored with species-typical song. Plotted is the change in mean frequency (in kHz) from the previous syllable(t-1) to the current syllable(t) (x axis) and the change from the current syllable(t) to the following syllable(t+1) (y axis). Arrows (in the corners) indicate the direction of change, with the first and second arrows of the pair indicating, respectively, change from syllable(t-1) to syllable(t) and change from syllable(t) to syllable(t+1). The significant negative relationship indicates that pitch alternated across syllables more than expected by chance.
(C) Birds tutored with randomized sequences of syllables also significantly alternated the mean frequency of syllables across the motif (r2 = 0.476; p = 0.0018; Monte Carlo simulations). Data for experimentally tutored birds are plotted the same way in (B).
Current Biology  , e4DOI: ( /j.cub )
Copyright © 2017 Elsevier Ltd Terms and Conditions