Here is Valerie’s entry for Task 3: Voice to Text:

Valerie Ireland – Task Three – Voice to Text

And here is my entry:

Task 3: Voice to Text

Both Valerie and I generated a text output that is remarkably similar from a distance. We both largely created something that is a massive wall of text, with little or no punctuation. While Valerie used Microsoft Word, I used the iPhone Notes application. We encountered similar issues with incorrect word choices made by the dictation engine and frequent grammar mistakes.

What I really enjoy about Valerie’s posting for Task 3 is the discussion about  context and culture specific limitations of the speech-to-text applications. Whereas I focused more so on the content of the mistakes being made by the software, Valerie discusses the impact of the mistakes. When trying to use dictation for something that is clearly outside the cultural context and understandings of the designers of the software, it is going to make many more mistakes than it would for someone from the same culture. For someone trying to discuss culturally sensitive or important topics, as Valerie mentions, this can be both a frustrating and painful experience. It can act to reinforce otherness and can imply that other cultures words and language are not important or valued. Even worse is Valerie’s experience of a word being translated into an obscenity by the software. These are not aspects that I considered in my Task 3, probably because I am from a similar culture to the creators of the software, and was discussing something that falls within that culture.

Software like this states that it works for ‘English’, but can it really be said to work for an entire language like English if it doesn’t functionally work for all types of English, for all speakers of English? It implies a ‘correct’ English, which is deeply problematic. It privileges a very specific cultural literacy, and excludes, sometimes violently, any others.