Figure 5.

Bottom-up (in orange) and top-down (in green) processes involved in the cortical tracking of the speech envelope. First, the speech stimulus enters the ears and its information is processed all the way through the first stage of the auditory cortex (arrow 1). Once this information reaches the cortical layers, the cortical speech tracking starts. Then two streams are initiated: 2a. a bottom-up auditory stream (see the 0–100 ms of the temporal model in Fig. 4); 2b. a bottom-up attentional stream, activating the occipital and the frontal areas. The parieto-occipital and frontal areas interact in the dorsal and ventral attention network (arrow 3). Finally, the frontal and parieto-occipital areas impact the later-latency auditory processes through top-down processing (see latencies between 150 and 300 ms in Fig. 4): the parieto-occipital (arrow 4; see the red circles in Fig. 1 in the Appendix) top-down control here preceeds the frontal (arrow 5 see the green circles in Fig. 1 in the Appendix) top-down control.