I am linking to Duncan Hamilton’s Voice to Text post.
Both Duncan and my posts explore the limitations of speech-to-text technology, but we approach the topic from different perspectives. Duncan’s analysis is thorough and analytical, identifying specific issues like punctuation errors, lack of structure, and frequent misinterpretations. He notes his surprise at the poor performance of Apple’s dictation feature, given his limited previous experience with speech-to-text technology. Duncan details how the software struggles with punctuation, resulting in run-on sentences and misplaced commas, and fails to maintain a logical structure, leading to text that feels disjointed. He also highlights instances where words were inaccurately transcribed, such as “tsukemen” being rendered as “superman,” and inconsistent capitalization throughout. Duncan argues that scripted text would provide better coherence, contrasting the informal nature of oral storytelling, which includes pauses and asides, with the structured format of written communication governed by grammar and punctuation rules. He draws on references like Ong’s and Gnanadesikan’s works to illustrate how writing transforms oral narratives into a more organized form, while acknowledging the challenges speech-to-text software faces in capturing the nuances of spoken language.
In contrast, my focus is more pragmatic, addressing key issues and proposing solutions to improve speech-to-text accuracy. I discuss how the time constraint during my recent use of the technology contributed to a lack of cohesion in the transcribed text. Like Duncan, I identify grammatical errors such as punctuation issues, capitalization inconsistencies, and structural deficiencies, including instances where the software misinterpreted words. Despite these shortcomings, I acknowledge that most words were correctly transcribed. I suggest that familiarity with specific speech-to-text commands could enhance punctuation and readability, advocating for a scripted approach to ensure a more organized and cohesive narrative. Additionally, I emphasize the flexibility and emotional depth of oral storytelling, which adapts with each retelling, contrasting it with the permanence of written stories. I also highlight the cultural significance of oral traditions, citing examples like First Nation educators who use storytelling as a means of preserving cultural heritage in the classroom.