Sunday, July 29, 2007


RSVP (Rapid Serial Visual Presentation) is an alternate way of presenting text. Instead of static pages of it as in a book, words are flashed for brief periods in front of you before the following words comes up. There are advantages and disadvantages to this method of visual madness.

Traditional reading assumes that text is static. Your eyes move over the text. This means there's a cognitive and biomechanical load on you: you have to move to each successive word and you also have to do so intelligently. But it also means you can re-read things and scan very easily.

RSVP reduces your biomechanical load: since the words don't move, your eyes don't have to move. It also reduces, arguably, cognitive load since you don't have to intelligently search for the next word or line. But you can't scan nearly as easily.

Apparently comprehension and speed with RSVP is, at best, comparable to traditional texts. Understandable, especially for those who haven't "trained" to read in RSVP style.

Anecdotally though, RSVP training is a great help in speed-reading traditional texts. This is because when you start getting used to it and start pushing up the presentation speed, your brain doesn't have enough time to mentally "vocalize" the words. Instead, you get used to entire words as graphical symbols. Then, when you go back to reading traditional texts, you skip the mental vocalization and instead look at whole words as symbols, a much faster way of deciphering a sea of letters and numbers.

Most current RSVP software seems to be just flashing word after word at a set speed. I wonder if there are ways to drastically improve that though. For one, many of the auxiliary words in the English language can be inferred from the context. Articles, for example, don't really need to exist for comprehension. Additionally, longer words take the brain longer to process.

I think RSVP speed can be tuned on a word-by-word basis with a good algorithm. For starters, the interval between one word and the next shouldn't be a constant. It should be a scaled constant, like time multiplied by length of a word. Shorter words get shorter presentation time while longer words get longer presentation time. The difference would be milliseconds but that's all the brain needs. Punctuation should be taken in to account as well. Often I'm looking at an RSVP presentation and the sentence confuses me because I missed seeing that little dot at the end of one word and a new idea comes flying at me without my realizing it.

Maybe I should write my own RSVP trainer and see where it gets me...

No comments: