Representation of Auditory Task Components and of Their Relationships in Primate Auditory Cortex

Abstract

The current study aimed to resolve some of the inconsistencies in the literature on which mental processes affect auditory cortical activity. To this end, we studied auditory cortical firing in four monkeys with different experience while they were involved in six conditions with different arrangements of the task components sound, motor action, and water reward. Firing rates changed most strongly when a sound-only condition was compared to a condition in which sound was paired with water. Additional smaller changes occurred in more complex conditions in which the monkeys received water for motor actions before or after sounds. Our findings suggest that auditory cortex is most strongly modulated by the subjects' level of arousal, thus by a psychological concept related to motor activity triggered by reinforcers and to readiness for operant behavior. Our findings also suggest that auditory cortex is involved in associative and emotional functions, but not in agency and cognitive effort.

Keywords: agency; classical conditioning; instrumental conditioning; monkey; sound (audio) processing.

FIGURE 1

Experimental design. The six rows show the six conditions used in this study which differed in the number and order of the task components S (sound), W (water), and M (motor response). The left column shows conditions in which only a tone was presented (tone alone) and the right column shows conditions in which a noise burst and a tone were presented (noise-tone). The first row depicts the S-condition in which a trial consisted only of sound, either a tone (yellow rectangle; 2.2 s duration) or a noise burst (pale yellow rectangle; 1.6 s duration) followed by the tone. The silent idle period between the sounds presented in consecutive trials varied between 3.5 and 7 s. In the W-condition, only water (W; blue rectangle) was delivered for 0.6 s in a trial and the idle period between consecutive trials varied between 5 and 10.3 s (for the sake of presentation the W-condition is shown in both columns). In the SW-condition, a trial consisted of sounds and water, the latter was delivered from 1.5 to 2.1 s after tone onset. The SW-condition was tested both in the Simple block (upper three conditions) and in the Task block (lower three conditions). In the MSW-condition, the monkeys themselves initiated a trial by bar touch. While holding the bar (black rectangle), sounds were presented and water was delivered. In the SMW-condition, a trial was externally initiated by a computer and started with sounds, in response to which the monkeys had to touch the bar at the earliest 0.3 but not later than 1.2 s after the onset of the tone. After holding the bar for 1.5 s, water was delivered. The monkeys had to release the bar and remain idle for 3.5–7 s before the next trial. In the MSMW-condition, the monkeys initiated a trial by bar touch. While holding the bar, sounds were presented. The monkeys had to release the bar from 0.3 to 1.2 s after the onset of the tone, which was rewarded with water 1.5 s after release.

FIGURE 2

Licking activity and auditory cortical firing rates differ between the three conditions of the simple block. In the first row, the plots show the licking activity of six experimental sessions in the instrumentally-trained monkey B for the S-condition (green), the W-condition (blue), and the SW-condition (black) in tone-alone trials (A) and in noise-tone trials (B). Note that the blue curves are identical in tone-alone and noise-tone trials already from 2.5 s before water delivery. Other conventions as in Figure 1. (C,D) Show the median population firing rates of 97 multiunits in the two instrumentally-trained monkeys. The shadings show the 95% range of the median. The rectangles and the short bars denote time intervals in which there was a significant difference in firing rates between the S and SW condition (gray) or between the W and SW condition (blue). Thin lines indicate p < 0.01; thick lines indicate p < 0.001. (E,F) Show the licking activity of fourteen sessions in the two passive monkeys. (G,H) Show the population firing rates of 75 multiunits in the two passive monkeys.

FIGURE 3

Population firing rates relative to the time of water delivery in the four conditions of the task block. Results are from 180 multiunits in the auditory cortex of two instrumentally-trained monkeys and are shown for the SW-condition (black solid curve), the MSW-condition (red solid curve), the SMW-condition (black dashed curve), and the MSMW-condition (red dashed curve) in tone-alone trials (A) and in noise-tone trials (B). Other conventions as in Figure 1. Note that in the SW-condition and MSW-condition, the firing peak ∼1.5 s before water delivery reflected the sound-evoked response. In the SMW-condition and MSMW-condition, this peak reflected the motor response to the tone.

FIGURE 4

Comparisons of population firing rates between pairs of conditions of the task block. The time is relative to the first sound in a trial. (A,B) SW versus MSW, (C,D) SMW versus MSMW, (E,F) SW versus SMW, (G,H) MSW versus MSMW. Other conventions as in Figures 1, 3. The dashed rectangles indicate the time interval in which the monkeys had to exhibit a motor response to the tone. The solid rectangles and the short bars denote time intervals in which there was a significant difference in firing rates between the two conditions shown in a panel, with thin lines indicating p < 0.01 and thick lines indicating p < 0.001.

FIGURE 5

Population firing rates of 180 multiunits in auditory cortex of two instrumentally-trained monkeys relative to bar touch. Conventions as in Figure 3. Data from tone-alone and noise-tone trials were combined. For the two self-initiated conditions, we disregarded the phasic sound responses to the first sound after the bar touch. The rectangles denote time intervals in which there was a significant difference in firing rates between the SMW and MSW condition (gray) or between the SMW and MSMW condition (blue).