The role of attention and emotion in message formulation

Abstract

The production of propositional language is unique and central to the human experience. Distinct from core spoken language skills (e.g., reading, naming, repeating), propositional language is voluntary, spontaneous, and novel to a context. Before producing propositional language, the speaker must conceptualise or formulate a message for expression. This preverbal stage of conceptualisation represents the interface between broader cognition and language. However, experimental studies addressing the relationship between cognition and spoken language in non-aphasic populations are scarce, and thus the interplay between language and cognition remains poorly understood. The aim of this thesis was to investigate the role of attention in propositional language production.This thesis includes four studies. The first study (Chapter 2) is an investigation of connected speech in mild stroke patients without aphasia (N = 18) and matched healthy controls (N = 21). We analysed the coherence and cohesion of connected speech samples, and assessed baseline cognitive skills including sustained attention, selective attention, and executive functions. Cohesion and coherence impairments were found in the stroke group relative to controls. Correlations between language measures and attention and executive skills revealed that better performance on the attention and executive tasks was associated with more cohesive and coherent speech. This pattern of results suggests that attention and executive cognition play a role at the stage of message formulation during propositional language production.The second study (Chapter 3) details an investigation of spontaneous speech and the frontal attentional process of ‘energization’ in a case series of five patients with progressive supranuclear palsy (PSP) and a matched healthy control group (N = 30). The process of energization refers to the initiation and sustaining of a response over time. We identified patterns of energization in spontaneous speech samples, word fluency, and an experimental concentration task. Four of the five patients with PSP had a response pattern clearly indicative of fluctuations in energization, whereas healthy controls were able to maintain consistent responding over time. Our results demonstrate a specific role of energization in spontaneous speech production.The third and fourth studies (Chapter 4) employed a novel experimental picture description task to investigate the role of attention and emotion in propositional language production. This Propositional Language Task required the verbal description of an emotional (positive or negative) picture, while ignoring an emotion word distractor (related vs. unrelated). Initially, this task was completed by a large group of healthy adults (N = 93), and the results revealed that participants were significantly slower to initiate speech in response to negative, compared to positive, emotional pictures. This suggests that the emotional content of the stimulus influences the conceptualisation of a message to be expressed.In a follow up study, the Propositional Language Task was completed by four patients with PSP and a healthy older control group (N = 26). The primary finding of the previous study was replicated in the sample of healthy older adults. An additional detailed analysis of the speech content was conducted to explore whether the distractor word interfered with the content of the propositional language produced. We found that distractor words appeared in the speech of the patients with PSP, despite being instructed to avoid the use of those words. Interestingly, when distractor words were unrelated to the picture, healthy controls incorporated the concept activated by the distractor word into their sentences. We interpreted the distractor word interference effects as relating to attentional failures during message formulation.Overall, this thesis provides evidence that attention and other aspects of cognition that modulate attention (e.g., emotion) play a role during the conceptualisation of a message. Our findings have important theoretical implications for models of spoken language production, and clinical implications for the treatment of disorders of propositional language.