Reading Like a Computer Reads

Spritz is a startup that wants to change how people read on small screens. The startup has created an app that feeds texts to readers one word at a time, arranged so that their eyes are not “forc[ed]…to spend time moving around the page.” You can see a demonstration in this video. The Spritz reading technique allows readers to see as many as 1,000 words per minute, and promises developers that it will make “streaming your content easy and more comfortable, especially on small displays.”

When writing about Spritz, one commonplace has been to note that the technique is a repackaging of rapid serial visual presentation, or RSVP, an approach to speed reading that is nearly 40 years old. Although Spritz claims novelty for their application of RSVP, the underlying idea has been around for some time.

Another commonplace in responses to Spritz is that reading at 1,000 wpm is not reading, exactly. Since RSVP is not entirely novel, we know how high wpm rates affect reading comprehension, and the speed reading makes reading comprehension much more difficult.

Although the marketing language for the service claims that it is adjusting reading technologies to our bodies (and, if nothing else, Spritz has been marketed brilliantly: I’m particularly fond of the descriptions on Spritz’s site of how much time we waste moving our eyes when we read; they remind me of the exaggerated problems that plague people in infomercials. I’m not sure that is a great way of thinking about what the software does. For instance, we already have devices that can look at words at an incredible rate of speed without comprehending them: computers. My Google search for “spritz” returned more than 3 million results in 0.31 seconds. Computers look for key words, detect the presence of words or other phrases, sort that information (a computer can quickly determine if, in a set of four documents, for example, the word “spritz” is contained in documents 1, 2 and 4, but not 3), all activities that one could reasonably imagine being encouraged by Spritz’s RSVP technique.

Why do humans need to read like this when computers do it already? Ian Bogost in The Atlantic calls the approach “reading to have read.” In other words, Spritz fulfills a need for a kind of reading fetishism, a desire to look at a lot of words that is disengaged from the need to understand what those words mean. Bogost argues that in this way, Spritz is a manifestation of of our need to consume rather than comprehend. Or, as he says, “Spritzing is reading to get it over with.”

Why does this matter to digital media and learning (DML) educators? I imagine that every schoolchild has had the experience of looking at a stack of books and wishing that there was a way they could magically scan through it all as quickly as Data does at the beginning of this clip. The desire to complete school readings just to get them over with is likely to always be with us, but educators shouldn’t think that applications like Spritz are going to be a magic panacea. Rather, we have to think of other ways to make reading engaging and vital to students’ learning. I believe that it is worthwhile to spend time discussing with students reading technologies, but we need to be clear-eyed about the affordances and outcomes of those technologies.

In that vein, we should also pay attention when technologies ask us to behave like computers. Interestingly, this is quite the opposite from how computers were originally designed. In “The Embodied Mind,” Varela, Thompson and Rosch note that digital computers were originally designed to mimic then-current scientific models of human cognition. The current vogue for massive arrays of computers arranged in neural networks is simply an updated version of this process: designing computer systems to mimic cognitive science’s models of the human mind. 

It is one thing to create technologies that we think will behave as we do; it is quite another to force ourselves to behave like our technologies. I’m not much of a moralist when it comes to technologies. I wouldn’t rule out the idea that there could be situations when reading like a computer would come in handy. However, we can only recognize those situations if we carefully examine our technologies and understand not only what they do but what they ask of us.

Baner image credit: Horia Varlan