A Better Way to Read

In the era of attention deficits, the new text will not be black and white.

An eye
Laszlo Balogh / Reuters

It’s rare to think about it, but moving your eyes back and forth over lines of text is one of the most commonly performed bodily motions.

When reading, your eyes go from word to word, left to right, one after another. When you hit the end of the line, your eyes make what’s called a return sweep. They go back to the left, to the beginning of the next line. During that sweep, we get a little time to process information. (Are you thinking about it now?)

That sweep is also where many of us mess up. We lose time. Most people don’t go all the way back to the first word, for example. We tend to land on the second or third word in a line, and then make another backwards movement to get to that first word. That’s inefficient.

Like any physical movement, they’re a matter of practice and coordination. The mechanics of getting text into one’s brain require skill apart from that involved in processing the meaning in that text. As with something like swimming or skateboarding, it’s a skill where most people can become proficient, but everyone’s capacity for speed and precision is not equal.

But there are ways to enhance our abilities.

To illustrate this, try reading this passage (from Larry McMurtry's All My Friends Are Going to Be Strangers):

Then try reading it like this:

The colors in this text are rendered in a precise and strategic way, designed to help people read quickly and accurately.

The most important feature is that each line begins with a different color than the line above or below. As Matthew Schneps, director of the Laboratory for Visual Learning at the Harvard-Smithsonian Center for Astrophysics, explained it to me, the color gradients also pull our eyes long from one character to the next—and then from the end of one line to the beginning of the next, minimizing any chance of skipping lines or making anything less than an optimally efficient word-to-word or line-to-line transition.

Improving the ease and accuracy of the return sweep is a promising idea for readers of all skill levels. And yet it’s one that’s gone largely ignored in the milieu of media technologies. Today many of us read primarily on screens—and we have for years—yet most platforms have focused on using technology to attempt to recreate text as it appears in books (or in newspapers or magazines), instead of trying to create an optimal reading experience.

The format—black text on white lines of 12 to 15 words of equal size—is a relic of the way that books were most easily printed on early printing presses. It persists today out of tradition, not because of some innate tendency of the human brain to process information in this way.

Meanwhile, people who aren’t especially skilled at intake of text in the traditional format are systematically penalized. People who don’t read well in this one particular way tend to fall behind scholastically early in life. They might be told they’re not as bright as other people, or at least come to assume it. They might even be diagnosed with ADHD, dyslexia, or a learning disability, or overlooked as academically mediocre.

“The book format was effective, but not for everyone,” said Schneps. “This is not just technology that could help people who are struggling with reading; this is technology that could help a lot of people.”

* * *

Our minds are not as uniform as our text. We all take in information in different ways. Some people read more quickly and retain more information when lines are shorter, or when fonts are bolder, or in different colors. The color-gradient pattern above is rendered by a product called BeeLine, developed by armchair linguist Nick Lum. He got the idea after learning about the Stroop Effect, the famous phenomenon where it becomes difficult to read words like “yellow” and “red” when they are written in different colors. Lum thought, “What if instead of screwing people up, we tried to use color in a way that helps people?”

After he won the Stanford Social Entrepreneurship and Dell Education startup competitions with the idea in 2014, Lum took to developing the technology full time. So far, the response from people tends to be binary: for some it’s a shrug, but for others, particularly people with dyslexias, it’s like turning on a light bulb. As Lum describes it, people tell him “Holy cow, this is how everybody else reads.”

The idea has been well received by reading experts, too.

“Most of the academic research is figuring out entirely what your eyes are going to do on one line,” said psychologist and Microsoft researcher Kevin Larson. “That has been such a challenge that it’s rare for anyone to pay much attention to what happens during that line return movement.”

At the University of Texas at Austin, Randolph Bias has studied the optimal length of lines of text for reading comprehension and speed. The two are generally at odds: Short lines make for a quick and accurate return (the movement is easier because it allows our eyes to take a greater downward angle than if the line were longer.) The downside is that because our brains process information during return sweeps, shorter lines don’t afford us that time. We also don’t get to take full advantage of peripheral vision – which is key. (He cites this as the problem with Spritz, the reading technology where single words rapidly flash before a reader.)

“Human beings have evolved to be able to take in 12 to 16 letters—at least—at a fixation,” Bias explained to me. This is known as our span of perception. “Just as some people are taller, some people can take in more at a time.”

So, long lines of text are better if you can minimize line-transition problems. And using color gradients seems to be one way to do that. Last year, optometrist Carole Hong did an initial, small study (not peer-reviewed, for the company) where she watched the eye movements of people as they read with BeeLine, and most of them skipped fewer lines–and experienced fewer backward saccades–than when reading in black text.

“BeeLine gives the reader a cue as to where that next line is – an excellent, unavoidable cue,” explained Bias. “I think that’s brilliant. It might be the first in a series of creative ways we break out of traditional formats of text – where we take full advantage of online capabilities instead of just plopping a book into a doc file.”

Beeline launched softly, accidentally, on the site Hacker News in 2013. At that point it was just a bookmarklet (the lamest version of a browser plugin), and the first version of the chrome extension. Even still, the concept captured enormous attention in the ADHD community on Tumblr.

“It wasn’t envisioned as an assistive technology,” said Lum, “but after the launch on Hacker News, we got all kinds of emails from people with ADHD. The folks on Tumblr are heavily in the ADHD and dyslexia camps.”

The color gradients might be helpful not just with return sweeps, but simply in keeping people’s attention – so they’re less likely to dart from tab to tab. Bias sees an important role for this technology in the era of waning attention spans. He’s 64 years old and describes himself as a “slow but good reader” who “can sometimes stay with something for a long time.” But in recent years, he’s sensed a decline in his attention, and has a feeling that this is a growing problem. “Can we multitask?” he asks, rhetorically. “The research, more and more, shows that we all suck at it.”

And as the definitions of paying attention and reading shift, narrowing the focus of technology like BeeLine to people with ADHD and dyslexia is missing the point. If changes in format like adding color gradients can help prevent loss of attention, then it necessarily helps with reading (retention and speed). Schneps says that what’s needed is a complete shift in thinking about normalcy—and why there is one default state for almost all text.

At Microsoft in Seattle, for example, Larson has been working for 19 years studying word recognition and reading acquisition. When he started, he recalls, very few people would read any long document on screen. If they got a long email, they would print it out. “But now,” he notes, “that would be an outrageous thing to do.”

The task now is to make digital reading better than reading in print. One of Larson’s passions has been working to prove that we take in words not as entities unto themselves (as some typographers have argued), but always by recognizing a word’s component letters, and then assembling them into words in our heads. His team also recently launched a new font that was designed for the best possible readability. Called Sitka, it went through a multistep, iterative design-test process. Each letter was changed and adjusted to maximize ease of reading – as opposed to most other fonts, which are made to mimic typefaces that existed in print media. “Times New Roman was designed to work very well with the technology of the era,” Laston explained. (I asked him if he has, then, created the most legible font in history. He said he “wouldn’t go that far.”)

It’s not the color itself that’s useful in Beeline, Larson believes, but the connection of the end of one line to the beginning of another. He suggests that for people who are colorblind or simply averse to the colored text, the same thing could be accomplished using bolder text or different fonts to draw people’s eyes along. “A feature like Beeline can encourage people to move from print to digital,” he said.

Indeed, many print media publications have struggled in recent years, but many persist—largely on the justification that people still like to sit back and hold a book or magazine in their hands. But if the digital reading experience continues to improve—adding value beyond mere portability—that could make analog reading ever more niche.

The magazine CNET recently partnered with BeeLine to published two fiction pieces online, where readers had the option to turn the colors on or off. Jeremy Toeman, vice president of products at CNET, was instantly attracted to the technology. “We’re all spending more and more time on screens, and getting tired of seeing black words on white backgrounds for hours and hours on end,” he said, “so anything that helps with that seems good.”

And the analytic data from the BeeLine trial at CNET suggests people like it. Readers who used the color gradients were more likely to read further down the page than people who didn’t—and more likely to read to the end. Depth that they read down on the page was almost half-again as much as other readers.’ In the world of digital media, this is not just a valuable metric, but possibly the most valuable. This is the type of authentic audience engagement that’s monetizable. Anyone can generate clicks; not many can compel people to read an entire article, when it is essentially competing for a person’s attention against every other thing that’s happening on the entire, enormous Internet.

As most of the world’s readers move toward reading primarily on mobile devices, smaller screens mean not just smaller text, but larger variations in ambience. The mobile reading environment is rarely as friendly as the desktop environment. You’re not parked in a quiet office, but commuting on a subway that’s accelerating and decelerating, and your very small text is moving in three dimensions, and your eyes need to track across and between lines. That’s much harder to do when you and the device are moving semi-independently.

“If we can get people to read more, in this era where everyone is swiping through everything in a millisecond,” Toeman said, “then I think that’s great technology.”

The other big opportunity for the technology is in educational settings. Later this year, BeeLine will be rolling out in libraries across California, as part of a licensing partnership. This is how Lum sees the company growing. The basic Google Chrome extension and iPhone app are free. But large-scale licensing deals with platforms and institutions like school systems could be more lucrative—and make the option accessible to people who wouldn’t otherwise think to try reading in color.

In early experiments, some students do seem to benefit from the color gradients. Last year, first-grade students in two general-education classrooms in San Bernardino, California, tried out Beeline, and many did better with comprehension tests afterword. “Because of my background in visual processing, I immediately wanted to check it out,” said Michael Dominguez, an applied behavioral analyst who directs the San Bernardino school district’s special education program. “Based on everything I know, it should work great.”

A suggestion for compulsive checkers