About ten years ago, my father, who never learned how to type, bought Dragon NaturallySpeaking software. I think it was version 2.0, but I'm not sure. My dad was hoping that the software would allow him to write faster and eliminate the frustration of having to hunt and peck at keys on the keyboard. As a high school student with many papers to write, I was also really excited about the opportunity to write by dictation. It sounded very easy and I was hoping that I could compose papers with my voice alone.
So, I plunged in--downloaded the software, "trained" the software to recognize my voice, and began writing/speaking. It was a nightmare. Each sentence I uttered was riddled with errors. I had to go back by hand--like a caveman!--and correct just about every word in the sentence before dictating the next one. I realized quickly that the software wasn't delivering accurate results. But I also realized that writing, for me, is not something that you just spit out in a fit of passion. For some reason, the advertising for the software had led me to believe that writing on screen was just like speaking aloud, but slower. I thought that writing was just something that poured out of you, and that doing it by dictation would make the pouring that much more efficient. But, this experience forced me to see that writing--for me anyway--is always slow and recursive. Now, writing recursively by voice--going back and forth--was nearly impossible. Plus, all of the extra errors made it that much more harder. If writing was slow before, now it was glacially slow.
The Dragon folks suggested that if errors were a major problem it was because the machine had not yet learned your voice and recommended that users take the training courses again. So I did. Basically this amounted to reading aloud several scripts on the screen. Then, I returned to writing by dictation and still found that I could not compose a comprehensible sentence. I gave up. My father, too, ran into similar difficulties and became very frustrated. He had paid quite a bit of money for the software and now it was going to sit on a shelf--never to be used again.
Now, it appears that Dragon is releasing version 10.0 of its NaturallySpeaking series. It claims to have 99% accuracy. I'm tempted to give it another shot, but I can't help being fearful. Based on my earlier experience, I've adopted a kind of "I'll believe it when I see it" approach. Besides, I learned the hard way what the advertisers will not tell you: to use this kind of software, you almost have to unlearn how to write by hand and relearn how to speak. It the (re/un)learning didn't pay off ten years ago, but maybe the technology is improved enough that it will pay off now.
In any case, if the 99% accuracy claim is true, then perhaps this is the beginning of a revolution in how we write. Honeycutt claims that "if word recognition accuracy increases to the point at which users can consistently produce clean drafts faster than thy can with keyborading, then voice recognition has a chance of becoming a widely adopted literacy tool in the corporate workplace" (315). I wonder if the technology might also become widely adopted by students, or are the differences between school-writing and workplace-writing too great? I also wonder if people will choose to buy the technology--even if it is better. Honeycutt seems to suggest that improvements in the technology will make it more popular. But many scholars have reminded us that whether a technology is seen as working or not working depends on the user, not the technology itself. It's possible that improvements in word recognition accuracy might not matter to people who like writing the old-fashioned way or who don't trust machines. There are a lot of such people. Still I wonder what the future holds: Will advertising reach and convince people that this is the next wave of writing? Will schools and/or businesses purchase the software and make writing by dictation part of the curriculum? Will writing by dictation supplant writing by hand?
It will be interesting to see what happens. Personally, think I might give the technology another shot--but not just yet.
Subscribe to:
Post Comments (Atom)
1 comment:
This post brought back memories. I recall getting software back in high school that would read aloud text on the screen. I also recall some of my students discovering similar software when I was a middle school teacher... that provided them with endless amusement (and me with occasional consternation).
This technology never really got beyond the point of amusement though, for me.
Thinking about it now, I find it hard to imagine listening to a 'computer' voice reading any lengthy text. I can't tell you how irritated I am by endless voice prompt hell during phone calls with corporate entities. Is it still useful, even if it's annoying? :) Apparently, if it allows the removal of the human touch, i.e. the removal of the human paycheck, then yep, I guess it's useful, for somebody at least.
Post a Comment