Inner speech is an internally perceptible speech-like experience which occurs in the absence of actual sound. For many, inner speech represents a part of the daily conscious experience, yet its cognitive and neural underpinnings remain unclear. Two issues in inner speech research are addressed in this thesis. First, no single cognitive model of inner speech generation fully explains the wide range of observed neural correlates. Second, the diverse phenomenology of inner speech and the potential for interactions with its generative mechanisms is often neglected. The aim of this thesis is to develop an integrated model of inner speech which explains heterogeneous findings and accounts for its diverse phenomenology. The predictive validity of this model is then tested using a range of experimental techniques. Chapter 2 used an fMRI meta-analysis to determine which brain regions are commonly seen across inner speech neuroimaging studies. Evidence for the involvement of both speech production and speech perception regions in inner speech was found, depending on the phenomenology of the elicited inner speech. Chapter 3 found that individuals with aphantasia, which is argued to stem from dysfunction of perceptual imagery mechanisms, exhibited deficits in inner speech which varied across phenomenological dimensions. Chapter 4 used a series of experiments adopting a dual-task paradigm to tax speech perceptual regions and speech production regions, respectively. Results demonstrated that motor and perceptual suppression efficacy vary depending on inner speech phenomenology, indicating a role for both mechanisms dependent on the type of inner speech involved. Chapter 5 used transcranial magnetic stimulation to suppress brain regions associated with speech production and speech perception, respectively. A trend towards stimulation of speech production regions suppressing all phenomenological varieties was observed, although this did not reach statistical significance. Overall, research findings varied depending on the phenomenology of the elicited inner speech - with the exception of Chapter 5. This supports our argument that inner speech should be interpreted as a dynamic and variable phenomena with varied generative mechanisms and neural correlates. However, the exact generative mechanisms and precisely how they interact with phenomenology could not be determined, with results across Chapters 2 - 5 not being sufficiently consistent with our integrated model, nor with existing models of inner speech.
|Date of Award||31 Dec 2023|
- The University of Manchester
|Supervisor||Gorana Pobric (Supervisor) & Bo Yao (Supervisor)|