| This plasticity of auditory processing has been observed at increasingly shorter timescales, from long term linguistic experience (Krishnan et al., 2005; Wong et al., 2007) to musical training (Musacchia G et al., 2007) to short-term training (Russo et al., 2005; Song et al., 2008), to context-dependent encoding (Chandrasekaran et al., 2009). The ability of the brainstem to modify its encoding in real-time is potentially attributable to feedback from the periphery, corticofugal pathways, connections to the limbic system, and entrainment to the stimulus. To investigate the mechanisms contributing to this real-time plasticity, it is necessary to consider the anatomical connectivity of the auditory system, and to study the modulation of the auditory responses as a function of context, both linguistic and nonlinguistic. To this end, we examined the modulatory effects of a preceding syllable on the auditory brainstem response to a syllable (linguistic context), and the modulatory effects of visually elicited emotion (non-linguistic context) on the cortical and brainstem responses to an unchanging auditory stimulus. In order to understand how emotional state influences a listener's physiological response to speech, subjects looked at emotion-evoking pictures while EEG evoked responses (ERPs) to an unchanging auditory stimulus were collected. This study found an effect of negative emotion as early as 20 ms after the stimulus onset, the earliest documented effect of emotion in auditory processing. At the cortical level, emotion was observed to directionally influence the evoked responses to sound in a valence-dependent manner and with a laterality effect consistent with previous studies. Moreover, the neurophysiological response was shown to correspond to the behavioral ratings given by the subjects. At the level of the brainstem, however, there was no valence-dependent effect of emotion, although a valence-independent suppression of background noise was observed. In order to characterize the effect of linguistic context on the auditory brainstem response and whether categorical perception played a role at this level of processing, speech tokens (a prototypical ba, an ambiguous ba, an ambiguous da, and a prototypical da) were synthesized such that a perceptual category boundary divided the acoustically ambiguous ba and the acoustically ambiguous da. The ambiguous ba was presented in two contexts: preceded by a ba prototype (within category: ba prototype and baba-primed ), and preceded by an ambiguous da, (across category: ambiguous da and bada-primed). We observed that the brainstem processing of a speech token is influenced by its phonemic context and that in the frequency domain, the response to an ambiguous speech token more closely resembles the response to the syllable it follows than to itself in a different context. Moreover, there was no effect of category boundary since the strength of the modulation did not depend on whether the pair was across category or within category. Considered with the context effects observed by Chandrasekaran et al., (Chandrasekaran et al., 2009) this effect may be mediated through a form of stimuli entrainment at the level of the brainstem. Taken together, these results indicated a previously undiscovered level of real-time modulation of the afferent auditory stream. |
