My dictation software arrived earlier this week. Responding to the rave reviews of such software from a colleague (who has used it to write his dissertation, among other things), I have decided to give MacSpeech Dictate a shot. At this point, I have used it for about ten minutes. It works pretty well, I think, but I want to step around the accuracy question–what everybody talks about–and get to three other issues. The first two came up when I mentioned the software to an online community of Grinnell people.
1. One Grinnell alum reported giving up on dictation software because it encourages the composition of short sentences. I had not considered this issue before setting up the software, but now I take the point. If I stick with dictation and want to keep my “writing” style the same in typing and dictation, I will need to suppress my instincts to make my speech syntactically simpler than my writing–or, and this may be the more interesting possibility, I will shift my writing style in the direction of my speaking style. In any case, I can imagine this issue bothering me. I can also imagine, however, that the software will encourage me to speak more deliberately in all situations and (finally) to cut down on my tendency to interrupt myself with “aa” and “um.”
2. One alum has already reported that she finds herself unnerved by remembering my paper comments and imagining me dictating rather than typing them to her. (She illustrates this sentiment with this image.) I bet other students would feel the same way if I call attention to the documents I dictate and those I type. Does this feeling represent a shift in our sense of the relative naturalness of orality, recording, typing, and images? Or did people who received dictated business letters back in Mad Men times find dictation a little creepy, too?
3. People who like dictation software (for reasons other than avoiding repetitive stress pain) generally cite productivity gains. I’m a reasonably good touch typist, though far from a great one. I can imagine dictating a little faster than I type, even accounting for the need to fix more errors in dictation. I think the main difference, however, is emotional: the dictation software expects me to say something, and if I do not, the software tries to figure out what I’m saying anyway. I wonder whether this is the key difference between typing and dictating (for people who like dictating): it switches the writer’s default setting from “don’t write” to “write.” As fans of behavioral psychology know well, switching the default can be a very big deal.
P.S. As I typed this post, I discovered that thinking about dictation software increases my awareness of how much time I spend correcting my typing mistakes. I wonder whether some people’s resistance to dictation stems in part from our tendency to notice the software’s mistakes more than we notice our own.
Nathan · October 24, 2009 at 10:26 pm
I haven’t followed the conversation preceding this post, but in response to #2, if good (i.e. human-equivalent) dictation software would spell the end of writing as the default mode of original composition, wouldn’t that just be a return to the way it’s always been (excluding the past 500 years or so)? I think of the biblical letters, for instance, which were exclusively dictated.
And re #1: Given your analysis, the next step in dictation software must be to mold the spoken word into written style (what I mean by human-equivalent). I suspect that something a good scribe/secretary does is to err on the side of the author’s writing style, beyond simply taking down a verbatim transcript. I’m sure stylistic changes would make up most of the work in polishing a dictated text, because speech and writing are simply different media. I can’t imagine Paul sitting down with the rough dictation of Philippians to take out the vocalized pauses and run-on sentences. (Not that Paul ever met a run-on sentence he didn’t like.)