While no one was looking…

Out of nowhere there seems to have popped up a lecture on youtube (well a summary of it, but a link to the full lecture) where a lecturer has developed (through a font program) shorthand converter. In the lecture, he puts in the text, and gets out a correctly (but strangely) written shorthand translation. Anyway check out the link:


(by michael_lisitsa for


10 comments Add yours
  1. That's so cool!

    I explored the site a bit. Very wonky — links behave differently sometimes.

    is a (sometimes) working version. Sometimes you need to reload the page between tests. Sometimes it gave me a Gregg translation, including some intermediate steps. Sometimes it gave me nothing.

    So cool!

    The …more… and samples links from that page give other things. Looks like he's aware of several different types of shorthand.

    I left a comment on the http://www.river-valley.tv page .


  2. I had the same problem with the web interface to his text2gregg program. I could only run it once, and then have to reload the URL to try another translation.

    But it's still very interesting. I couldn't find a download link or source code on his site, I hope he makes it available.

    From his page there is a link to a one-page PDF which briefly describes the program: http://www3.rz.tu-clausthal.de/~rzsjs/steno/TeX/text2Gregg.pdf

    It seems to use a dictionary for common briefs and phrases, and depends on the Unisyn lexicon (http://www.cstr.ed.ac.uk/projects/unisyn/) to infer the pronunciation of the remaining words.

    I haven't watched the video yet, but I'm downloading it now so I can when I have time. Thanks, Michael, for pointing it out!

  3. Very cool.   I tried it, but it wouldn't transliterate the word "himself". Just left a blank.   And he doesn't seem to understand the difference between Anny (the page with the alphabetic symbols), Simplified (the text shown) and Centennial (the dictionary used).   sidhe

  4. I think Dr Gregg might think it an interesting exercise, but not very important. After all, the point of shorthand in the old days was to get words onto paper.   He'd have been much more interested in a computer that could OCR handwritten Gregg into text.   I'm sure it could be done. If a computer can scan and identify faces…   sidhe

  5. Okay, this may seem like a stupid question but here goes:  isn't this all sorta backwards?  Using a computer to take the written word and turn it into shorthand seems to be counterintuitive.  Typically the process is the other way around.  Is there a realtime application that would make this useful?   

  6. I realize this is an old thread, still interesting. Besides being a very cool programming project, whether it's backwards or not depends on a person's intended purpose. I'm working on my reading fluency, and this could be handy.

Leave a Reply