woops, sorry wrong link. here she is: https://www.youtube.com/watch?v=yJptrlCVDHI
Why should that be so (other than in extreme cases where pronunciation is way off what is laid down in the software - which is becoming increasingly sophisticated in its ability to interpret different accents and dialects)?
Fundamentally, what the computer is doing is matching printable text from a dictionary to electronic waveforms created by sound. It’s not clear to me why that should require understanding of language (except, as I say, in ‘edge’ circumstances, which the software presumably tries to minimise in a growing majority of cases by greater precision in the matching process - hence the year-by-year reduction in errors). Of course, understanding language would help, as it does with hearing-impaired humans - providing context and so on - but in most cases isn’t needed.
Creative writing, good creative writing, I think is an arduous process for a computer to have to decipher. Consider for example George R.R. Martin or Cormac McCarthy’s writings. Do you really think with all the extremely obscure or thousands of made up words both these author’s use, that writing their books via speech recognition is really all that feasible?
There is something very spiritual about the act of writing that I think voice recognition just doesn’t seem to capture. Maybe it’s the fact that the act of writing is primarily a silent act, that makes speech recognition seem a tad tangential.
Still, for people with disabilities, I think speech recognition is a valid and in many cases the only viable option.
I’m hoping in the future, that speech recognition improves. But as it stands, I think the technology is just not there yet.
Sorry, I should clarify. Cormac McCarthy = really obscure words; George Martin = really made up words; writing science fiction and fantasy = lots of made up words, ideas, etc; writing other forms literature like poetry = makes the act of writing very pleasurable, something that can’t easily be translated with speech recognition.
Have you ever actually used speech recognition software? “Teaching” the computer obscure or invented vocabulary is trivially easy and has been routine for a decade or more: you just say “this waveform matches this combination of letters.”
For accurate dictation, the computer has no need to know anything about meaning or grammar, and in fact it will quite happily transcribe any random series of words that you happen to string together.
Which is why entire novels have been written using voice recognition software on consumer PCs, but it took a supercomputer to win at Jeopardy.
Katherine
“Writing” was an oral act before it was a physical one. All of our earliest literature comes from oral traditions: epic poetry, theatre, storytelling.
But, again, computers and tablets and pens and voice are all just tools that facilitate or impede the mind of the writer.
Katherine
I’ve been using Dragon on PC since some earlier versions, and now I use Dragon 3 Italian on Mac. I agree that voice recognition is really mature today (and I await to see Dragon 4, not yet available in Italian). I can dictate very fast, speaking in a natural way, with a minimal amount of errors. In fact, I dictated my 120-pages dissertation, and it worked great.
But yes, I don’t feel still at home at creating narrative dictating. But I know this can be done, as several writers have done with their secretaries, and I would like to learn it much better than I can do now.
Paolo
I live in the countryside, and in a country that invested very little in public transportation. So, I always move with my car, and the keyboard can quietly sit in the car. So, I don’t feel as if I had to “carry” it around. It’s there with me.
I use my iPhone for capturing ideas everywhere and everytime. A few days ago I was at a mountain resort, walking along a spring, and had to tap an idea on my phone. I activated cell data, and my idea was immediately sent to Dropbox (on the cloud, the iPads and my Macs), ready for my return at home.
Sometimes I cannot (or want to) type, and record my idea. When at home, I sync my phone and listen to the voice note in iTunes.
I would say that what you would like to see in an app already exist. At least, I think I’ve been using it for a few years.
Paolo
Yes, I have tried using Speech Recognition as I suffered from severe De Quervain’s tenosynovitis for several years and couldn’t type without pain. But no matter how hard I tried, the computer kept on making mistakes whenever I tried to dictate most of the time and it was all an altogether extremely frustrating experience.
There were many occasions where I felt, jeez, if the computer actually tried to understand the meaning behind what I was saying, the content and the context, it’d probably make less mistakes. Pressing the backspace key to delete things, highlighting words, and the whole process of editing just seemed much easier when I was typing things out. The whole “scratch that” schemata felt largely unnatural…and teaching the computer new words every time I made one up was an extremely annoying (and laborious) task.
Yes, this point about writing coming from oral traditions is very true. John Milton orated Paradise Lost completely blind. And there were certainly no computers during Shakespeare’s heyday.
But typing has become such an integral aspect of ‘writing’ that it really is difficult to think of it without it. Yes, it’s possible. But my brain, for whatever reason seems hardwired, to think at the same pace that I type. I just don’t have the same connection when I talk. Or when I handwrite. I’m the progeny of the digital age I guess…
In the end I think there are lots of issues with speech recognition. My hope is that it gets better with new computer processors and programming and such…
I like that. Well, I’m a Child of the Sixties, which carries its own costs and benefits…
I suspect that there’s not a great deal of disagreement between us. I believe in drafting and, where necessary, re-drafting, which is a reason that I’m a fan of Scrivener. Dictation for me is just a tool to use to produce one of those drafts - but certainly not the first (unless the final document is to be no longer than three or four paragraphs). At some point in the creation of a longer work, it ought - in my view - to pass through the fingers (either by typing or hand-writing or both).
Of course, as posters have mentioned above in this thread, there have been some published authors who do compose “on the hoof” right from the get-go. Perhaps the most famous of these in the UK was Dame Barbara Cartland, who wrote more than 700 published books and several plays in her lifetime, most of them historical romances, by dictating to secretaries. You can read about her here: http://en.wikipedia.org/wiki/Barbara_Cartland. In her case, “on the hoof” meant almost invariably “on the chaise-longue”. (How many Scrivenistas today compose reclined on their chaises-longues? The Wikipedia section on Dame Barbara’s marriage and relationships is also quite amusing, bearing in mind that she was a romantic novelist.)
700! Holy crap. Here I am trying to write 1 book and even that I find a struggle…
Well, yes, she wrote more than 700 books. However, having read these romance novels as a teenager, I can tell you that she used a formula and the only real difference are the names, eras, and countries, so to speak. And she never went far beyond “they finally kissed at the altar. They lived happily ever after. The End.” Look for her books in the used bookstores and you’ll see what I mean.
Here’s a partial sample of a book that I found at Smashwords.
yeah, there is quantity and then there’s quality…
then again, writing ‘pulp fiction’ is difficult in its own way. everyone looks down on pulp fiction writers, but I tried my hand at writing something really pulpy sometime ago and I just couldn’t do it…
‘Penny dreadfuls’…are not that dreadful, in my view. they require a specific tenacity to recognize the hegemonic issues surrounding modern life, and write in a way that’s, simply put, profitable.
Stephen King might be ‘pulpy’…but pulp fiction is an art in itself… and if we should gauge one type of writer, we should gauge him or her in reference to other writers of the same kind.
People can say all they want about Daniel Steele, but she’s a millionaire…and I’m well…not.
Thing that worries me, will iScriv interfere with iPhotoShop on my iWatch?
ps
don’t know how i’d write on an apple watch…but I guess, if it can be done, it will eventually be done.
Ulysses for iOS is coming out already! In addition, this comment is not precise: Daedalus syncs also via Apple iCloud and Dropbox in a really effective way. I know this because I also use it.
I have also seen the Storyist combination Mac-iOS and seems very improved.
I am sorry to say that ‘Scrivener for iOS’ delay risks to put the whole software in danger!
I never said that Daedalus does not sync via iCloud. But if you use it, you should have already discovered that Daedalus only supports a reduced set of Ulyesses’ features. This is well documented in the FAQs on Soulmen’s web site (and can easily be verified with a few tests with multiple folder levels, images, and document notes).
As for Ulysses for iPad coming, I know it is in beta. I don’t know the status of Scrivener for iPad. Who knows which of the two will appear sooner?
Paolo
I received a mail from Soulman stating that Ulysses for iPad will be out in January 2015. I don’t know if this will be the case, but I doubt that Scrivener for iPad will be ready for that time. The last blog-post on this issue seems like the too many letters we read on the issue during the last two years, full of details on the specific reasons why Scrivener for iPad is so difficult to accomplish. As I already posted many times, priority has been given to other things (such as the Windows version, for a good commercial reason I am sure) and the iOS project has been put aside. This is the simple reason.
I would be really happy if my forecast would be wrong! W Scrivener, anyway!