While it is a common belief that the proofreading of transcribed audio is an unambiguously black-and-white affair of matching spoken words to text, there are many challenges and grey areas inherent in the task.
In the first instance, a proofreader receives an audio recording of a meeting, interview, business forum, et cetera, and a corresponding Word document containing its transcription. The task is to listen carefully to the audio recording and to check the Word document for accurate transcription, while simultaneously reviewing spelling, punctuation and document formatting. This task is in itself quite challenging, particularly as the human brain is wired to give preference to visual stimuli over that of other senses, and therefore to confirm the text we see, even when there are anomalies in sound. This means that when a spoken word or phrase sounds slightly different to its transcribed counterpart, we are predisposed to believe our eyes, and accept the written word before us. For this reason, proofreading requires an almost superhuman level of attention and concentration at all times, along with acute listening skills. To add to these challenges, the proofreader often works in the dark, lacking basic information about their audio file, including the number and names of speakers (as well as correct spelling). He or she may also be unfamiliar with terminology or acronyms specific to the project, organisation or industry from which the file originated. Typically, the proofreader begins with little contextual information or detail about the file or its source, making it necessary to research terms and even speakers on the Internet to confirm details.
Another significant challenge for the proofreader is his or her unfamiliarity with individual speakers or the sound of their voices, making it hard to identify or distinguish between speakers – particularly those with similar accents or voices. Often speakers talk over one another, and it becomes difficult or impossible to decipher what was said. When short utterances of one or two words are spoken, it is often problematic to identify the speaker, further increasing the complexity of the task. Poor audio quality and background noise, foreign speakers with accents and atypical English language phraseology, and quiet or mumbling speakers only heighten the challenge. Unfamiliar place names, tribes, languages or cultures in remote locations are sometimes indistinctly referred to and impossible to Google. Yet another challenge arises from fast speakers who are hard to keep up with! In combination, these factors can create multiple shades of grey.
As part of our process, words or phrases that cannot be deciphered by the transcriptionist are time-stamped. The proofreader’s role is to attempt to resolve these, slowing down the process further, as several repetitions of the audio are often required to confirm any changes to the text. Regardless of whether the word or phrase is resolved, cross-referencing is required between the time shown on the audio recording and its position in the Word document, to ensure correct stamping.
All in all, the work of both transcriptionists and proofreaders is a great deal more complex than the black-and-white text of their documents would suggest, requiring advanced English language skills, excellent hearing, and intense concentration and attention to detail. It is certainly the most challenging and time-pressured work I have ever done.