While it is pretty interesting for quick reading of small snippets of text it begins to fail when trying to build up a comprehension map of a large body of information. Especially if that information is of a highly technical nature. The lack of comprehension is compounded; the more unfamiliar you are with the concepts at hand.
One of the things that we do without thinking about it is make associations between things we know and things we don't know when reading. This allows us to provide a framework for the words and phrases we don't understand using the ones that we do, thus pulling the meaning out of the context. Spritz and by extension Squirt does not allow you to create these links because you don't retain the full breadth of the context for each word as you are reading them, and with no built in method for jumping back or forth through the information being presented it fails to allow these bridges to be built, and this lowers comprehension.
Now take something like this and implement an interface that tracks the users eye movement via a webcam to allow the flow of words to jump backwards and forwards through the surrounding information might work. But you would have to do some pretty impressive eye tracking and do it all with as little delay as possible to make the flow smooth, no matter what direction it goes. That said jumping back in the information would probably be jarring if the text just starts to flow in reverse. So one would probably have to jump back with some kind of visual cue to a pointer further back in the text and then continue reading forward again.
Also I'd love to see this implemented where it shows a sentence at a time. I'd be interested in the overall improvements to retention and understanding that would be gained with a reduction in potential top speed per word read. How drastically does speed fall off when you do this?
One of the things that we do without thinking about it is make associations between things we know and things we don't know when reading. This allows us to provide a framework for the words and phrases we don't understand using the ones that we do, thus pulling the meaning out of the context. Spritz and by extension Squirt does not allow you to create these links because you don't retain the full breadth of the context for each word as you are reading them, and with no built in method for jumping back or forth through the information being presented it fails to allow these bridges to be built, and this lowers comprehension.
Now take something like this and implement an interface that tracks the users eye movement via a webcam to allow the flow of words to jump backwards and forwards through the surrounding information might work. But you would have to do some pretty impressive eye tracking and do it all with as little delay as possible to make the flow smooth, no matter what direction it goes. That said jumping back in the information would probably be jarring if the text just starts to flow in reverse. So one would probably have to jump back with some kind of visual cue to a pointer further back in the text and then continue reading forward again.
Also I'd love to see this implemented where it shows a sentence at a time. I'd be interested in the overall improvements to retention and understanding that would be gained with a reduction in potential top speed per word read. How drastically does speed fall off when you do this?