Tuesday, May 11, 2010

Is Voice Better than Text?

I recently read about Nuance Communication’s Dragon Dictation app for the iPhone and iPod Touch. Nuance has created this app version of its remarkable Dragon Naturally Speaking software which allows you to speak into a microphone while Dragon types the text for you. Distance learning professionals have loved Dragon Naturally Speaking and comparable software for years. Many distance learning professionals will also gravitate toward to Dragon Dictation app which adds greater mobility to their ability to work at a distance. (This app is currently free in the US and Canada but if it becomes popular the cost will undoubtedly rise.)


Other Voice to Text Technologies

Of course, Dragon Naturally Speaking has not been the only technology to transcribe voice to text. Microsoft has long built this capability into its operating systems. However, it has not been something that they have marketed as value added, perhaps because it has been woefully inferior to Dragon until recently. Although Dragon Naturally speaking is still in my opinion the best, Microsoft has recently put some pretty good speech recognition software in Windows 7. It starts up easily and works a million times better than their previous versions although Dragon is still better.

Problems

With the app and speech recognition, professionals have gained powerful new tools to help them conduct distance learning courses which are largely text-based. However, does the use of the tools translate into better courses or better service to the students? From my own experiences the answer seems mixed.

One problem is over-reliance on the software. Dragon boasts that it is 99% accurate—and it is amazingly accurate—but there are drawbacks. Dragon learns to recognize your individual speech patterns as you use it and that takes time for the accuracy to get really good. Even so it makes mistakes and herein is a significant difficulty. You need to proof-read your own text from Dragon which is tough for lengthy passages where Dragon got 99% of it right but there are hidden sections that Dragon bobbled but with real words.

Years ago when I first installed Dragon I felt liberated from my slow typing speed. Now I could “type” at the speed of thought. For my online students this meant that I could answer their questions with more detailed, more nuanced answers. In actuality, many of my answers because less succinct. The reading involved in an online class already takes a good deal of time. Long instructor passages take even more time. Lengthy answers from instructors can confuse those who are not strong readers.

It has been my experience and I have heard from others that there is a fair amount of mental processing while typing with the keyboard. For most people, speech may happen to rapidly for this to occur. Speech is also perceived as relatively informal, a fact which we easily tolerate when speaking to someone, but becomes written speech significantly increases the expectation that there will be a more formal structure to the content and that more thought will be embodied in the construction of the content.

Why transcribe at all?

Sometimes I wonder why we bother transcribing at all. Would it not be better if we are speaking anyway to simply record the lecture and post the audio? This would help to students who may have some kinds of learning disabilities. This could help students who have differences in learning styles. This would enable students to listen to the content while mobile such as on an iPod or other mp3 device, which fits with the busy way our world works now.

Some of my colleagues argue that making the text available to students is important still to promote textual literacy, that it is still important for students to master written communication. Yet the same colleagues do not know what to make of the current texting phenomenon, which has students producing more text-based communications than ever.

Perhaps programs like Dragon and the new Dragon app could make a huge impact if adopted for texting. In fact, it could help us solve some of the issues we are fighting in education such as sentences in all lower-case with no punctuation and myriad acronyms. Teens, and all of us, could type as fast as the thoughts occur. This would remove the need for large QWERTY keyboards on phones. The question is would it be adopted because again we do not understand why texting is so popular. When we do we will be able to address the educational problems and move more decisively toward changes in the way students learn and the way we provide them information online.

1 comment:

  1. could be good if you are writing a book. you could edit the mistakes and go on!

    ReplyDelete