Rapid Responses to Auditory Frequency Change, Its Magnitude and Direction –
from Brain to Action

Amos David Boasson,1 Gal Vishne1, Leon Deouell1,2, Roni Granot3
1 Edmond and Lily Safra Center for Brain Sciences, The Hebrew University of Jerusalem, Israel
2 Department of Psychology, Hebrew University of Jerusalem, Israel
3 Department of Musicology, Hebrew University of Jerusalem, Israel
1 agboas@gmail.com, 2 gal.vishne@gmail.com, 3 leon.deouell@mail.huji.ac.il, 4 ronigra@gmail.com

Background
Auditory frequency change [FC] may convey to perceivers potentially crucial environmental cues, entailing immediate behavioral responses. Our research-line explores swift (albeit at times covert) effects of FC and its parameters on human motor action, and the junctions where this auditory information may be relayed into motor commands.
Previously (Boasson & Granot, 2019) we employed a finger-tapping task, monitoring negative mean asynchrony, muscle-action, and finger acceleration. Isochronous beeps (rate 4 Hz) were set in sequences presenting task-irrelevant increments/ decrements, in frequency/intensity. FC elicited augmented asynchrony, revealing 'melodic' asymmetry: yet more enhanced 'tap-earliness' followed Rise (thus no mere surprise effect). Physiological data detected Rise's effect-onset at ~160 ms post-FC; Fall impacted action significantly later. Diverging response-patterns to FC vs Intensity Change suggested domain-specificity.
Aims
Focusing on auditory-to-motor mediation, the present study explored FC effects on both behavioral and neural activity at a motor task evoked in response to FCs presented at random time intervals (instead of activity pre-planned by synchronizing to a regularly-paced auditory input, as in the tapping paradigm).
Method
In a demanding reaction-time [RT] task, 32 musicians heard rapid sequences of single-frequency pure-tone beeps (beep duration 86 ms) at semi-randomized stimulus-onset asynchronies (average 8 Hz). At semi-randomized inter-trial durations (average 1 Hz) FCs occurred; their magnitude and direction were manipulated (using seven tone-frequencies, range 231-1219 Hz; frequency step 'unit' 480 cents; single to quadruple steps; equal loudness 75 dB). Sixty-six sequences of intermixed lengths (36|45|54 trials, total 2970 trials) were grouped into eleven blocks.
Single and double steps were at focus. There was equal probability of departure by rising/falling single/double steps from the three center frequencies, and of arrival by rising/falling single/bigger steps to the mid-frequency (530.5 Hz). The design enabled distinguishing effects of FC direction and FC magnitude from those of absolute frequency.
A button-press was required upon perceiving any FC. Utmost rapidity and false-start avoidance were encouraged and monetarily rewarded.
We recorded RTs, finger acceleration, brain and brainstem activity (EEG), and muscle activity (EMG). Analyzed components included ABR (Wave-V), Na, Pa, Nb, P50, N100, and FFR. EMG and accelerometer data analysis included time-binning, cluster permutations and single-trial. For RT analysis, data were trimmed asymmetrically (15%, adaptive robust method).
Results
RTs were rapid (mean 203 ms, 92% valid responses), being significantly affected by FC magnitude and direction: shorter to large (vs small) steps [ηp2=.90], and shorter to melodic fall (vs rise) [ηp2=.35].
Mean muscle response-onset was identified at <70 ms post-trial-onset [!]. Significant differences in response-onset between the different FC types, consistent with RTs, were observed. FC magnitude shed stronger and earlier effects than FC direction. Finger acceleration results corroborated the EMG results.
ERP data revealed decoding of FC, its magnitude, and its direction, already at brainstem and through the following components. FC (vs non-FC) tones elicited enhanced peaks at all components (brainstem to N100), enhanced mean amplitudes (Na to N100), and earlier peak latencies (Na, Nb). Regarding FC magnitude, bigger (vs single) steps yielded a higher brainstem peak, earlier Na+Nb peaks, smaller Nb, enhanced P50, and earlier N100. Regarding FC direction, rise (vs fall) yielded earlier peaks but lower mean amplitude at brainstem, earlier Pa peak, and (marginally) enhanced Pa and P50; however, fall elicited earlier and enhanced N100. The effects of FC magnitude on Na & Pa latencies and on EMG onset latencies were (marginally) correlated.
Our novel findings regarding early detection of FC magnitude and direction from brainstem onwards extend previous knowledge which mostly marked N100 as the earliest distinguishing component.
Our findings regarding FC magnitude effect on RT (big→early) align with previous literature; however, our findings regarding FC direction (fall→early) do not. This latter divergence may stem from our subjects' stressing task: aversive conditions may render falling FCs threatening, eliciting earlier action.
Super-rapid RTs necessitate early mediation of auditory input, via 'decisions', into motor commands. Accounting for brain-muscle neural transmission time (estimated ~35 ms) suggests motor command issuance at ~35 ms post-trial-onset, within the Pa timespan. Indeed, Na and Pa, reflecting processing-levels between MGB and A1 cortex, did show latency correlation to EMG-onset. Interestingly, ascending pathways to MGB involve axons branching 'downwards' to brainstem/spinal-cord, suggesting possible mediation to motor routes.
Conclusions
Detailed FC detection early up the auditory pathways elicited (in musicians) ultra-early action. Hence, early ‘decisions’, based on finely-processed information, yield motor commands at early primary auditory cortex levels, perhaps even sub-cortically. Upcoming, testing non-musicians could further our insight.
Bewildering our concept of consciousness threshold, bottom-up and top-down processes meet at 'low' levels, activating auditory-to-motor relays. Enabling rapid responses to FC information, such primal routes may act regularly, though covertly, possibly forming motoric 'low-level' facets of auditory perception, both generally and in realms focusing on distilled FC, as Music.
References
Boasson, A., & Granot, R. (2019). Short Latency Effects of Auditory Frequency Change on Human Motor Behavior. Auditory Perception & Cognition, 2(1-2), 98-128.
Keywords: Auditory frequency change, perception-action, auditory-motor relay, embodied music cognition, auditory processing