Date of Graduation


Document Type


Degree Type



School of Medicine



Committee Chair

James W. Lewis

Committee Co-Chair

Albert S. Berrebi

Committee Member

David W. Graham

Committee Member

Marc W. Haut

Committee Member

George A. Spirou.


The human auditory system is able to rapidly process incoming acoustic information, actively filtering, categorizing, or suppressing different elements of the incoming acoustic stream. Vocalizations produced by other humans (conspecifics) likely represent the most ethologically-relevant sounds encountered by hearing individuals. Subtle acoustic characteristics of these vocalizations aid in determining the identity, emotional state, health, intent, etc. of the producer. The ability to assess vocalizations is likely subserved by a specialized network of structures and functional connections that are optimized for this stimulus class. Early elements of this network would show sensitivity to the most basic acoustic features of these sounds; later elements may show categorically-selective response patterns that represent high-level semantic organization of different classes of vocalizations. A combination of functional magnetic resonance imaging and electrophysiological studies were performed to investigate and describe some of the earlier and later stage mechanisms of conspecific vocalization processing in human auditory cortices. Using fMRI, cortical representations of harmonic signal content were found along the middle superior temporal gyri between primary auditory cortices along Heschl's gyri and the superior temporal sulci, higher-order auditory regions. Additionally, electrophysiological findings also demonstrated a parametric response profile to harmonic signal content. Utilizing a novel class of vocalizations, human-mimicked versions of animal vocalizations, we demonstrated the presence of a left-lateralized cortical vocalization processing hierarchy to conspecific vocalizations, contrary to previous findings describing similar bilateral networks. This hierarchy originated near primary auditory cortices and was further supported by auditory evoked potential data that suggests differential temporal processing dynamics of conspecific human vocalizations versus those produced by other species. Taken together, these results suggest that there are auditory cortical networks that are highly optimized for processing utterances produced by the human vocal tract. Understanding the function and structure of these networks will be critical for advancing the development of novel communicative therapies and the design of future assistive hearing devices.