Can we use coloured text to speed up reading?

According to Wikipedia:

German gothic text
German gothic text

In one common form of synesthesia, known as grapheme → colour synesthesia or colour-graphemic synesthesia, letters or numbers are perceived as inherently coloured.

My thought is whether artificially helping people associate letters with colours, can increase the speed of reading, specifically, in ‘specially prepared’ texts read from a computer screen, but I’m also interested in whether it might persist away from there.

Why?

Most people read by pattern matching the first two letters (ish) of a word – it’s how the neolism Typoglycemia works: for how you can largely understand:

“Amzanig huh? Yaeh and you awlyas thguoht slpeling was ipmorantt.”

Often though, we’re reading text on a computer, that the computer can help us with. That is to say, a webpage, an email, an ebook. Computer manufacturers spend a lot of time developing typefaces that are easy to read, and hard to confuse the letters of (Google even came up with an entire typeface for Android).

But it’s not always possible or desirable to read things in typefaces that are different from the original,

My theory is that the brain can probably increase its word-based pattern matching skills, by assigning each letter of the alphabet a shade of colour.

This might sound like the most ridiculous thing you’ve ever heard or seen, but if you consider that programmers used syntax highlighting to quickly derive extra meaning from great blocks of text – it seems more reasonable that there might be some way of using colours to improve up pattern matching when reading words.

So take a look at this prototype colourphabet I just threw together.

qwertyuiop

asdfghjkl

zxcvbnm

so

The quick brown fox jumps over the lazy dog

You’re not the only one thinking “this is much harder to read than ‘The quick brown fox jumps over the lazy dog'” and I don’t propose that I’ve solved this, or got anything more than the start of a stupid-sounding idea.

I wonder if it might better apply with different colour pallets, or perhaps colour pallets applied to different words in sentences – perhaps based on adjective/verb/etc… or something else?

How would you improve it? I’d love to hear your ideas!

One thought on “Can we use coloured text to speed up reading?

  1. Hi Tim! This is super duper interesting stuff. You’ve touched upon a major area of study in linguistics, specifically psycholinguistics, which delves into how we process language in the brain (among many other things). What you’re kind of wanting here is a sort of reverse Stroop effect (https://en.wikipedia.org/wiki/Stroop_effect), where the color of the text _aids_ the reader in getting to the correct meaning/understanding of the visual cue faster than it would in unmarked text. This can be tested and measured (and has been in lots of ways already)!

    What you’ve got currently with the letters mapping onto colors actually _adds_ interference/noise and _slows_ processing time. This is for many reasons, but the one that jumps out to me instantly is that by mapping colors onto _letters_ instead of units that are more meaningful for processing language, such as words or morphemes (the meaningful parts that make up words), the viewer is forced to make sense of the letters in isolation, distracting from whole-word comprehension (this ties into proven theories about how we actually read). It’s in effect jarring because we get extra information that doesn’t add redundancy or help to what we’re processing. For an example of what I mean by redundancy, think of the sentence,

    “She is a girl.”

    In this sentence, there are two places where gender is conveyed (the pronoun ‘she’ and the meaning of ‘girl’). There is also three places where number is conveyed (singular ‘she’, singular verb ‘is’ and singular ‘girl’). There are other, more complex redundancies in this sentence, but the idea here is that if we miss any of the ‘signal’ (e.g., due to weird font, bad vision, low light, whatever), we can recoup that loss because some of the information about what’s going on there is reinforced/repeated throughout the entire sentence. (It’s also part of how we can sort of make sense of what was said when we’re in a noisy bar and somebody is yelling at us).

    Additionally, we can process that sentence above faster when there’s redundancy because our brain can make logical guesses about what comes next — after reading ‘She is a’ your brain has an idea that ‘girl’ is more likely to come next (as opposed to, say, ‘giraffe’, and you don’t even need to fully process/”see” the word, visually or otherwise, before your brain knows that ‘girl’ is correct! (this where Typoglycemia enters into the picture btw) It’s super weird and complicated how our brains do this on a variety of levels all at once, but they do.

    To go back to your idea, you could possibly get this to work if you **trained** people to associate colors with something meaningful toward processing, e.g., all subjects are red, all verbs are orange, and all objects are yellow. Then the sentence ‘She is a girl’ would be red-orange-yellow, which maps what we already know about the color spectrum (order of the rainbow) with common (subject-verb-object) sentence patterns in English. This gets confounded really fast though, even with training through lots of examples. And synesthetes would probably actually be slower at processing this, because they have other mappings between color and language already, adding extra noise. Also worthy of noting, not all synesthetes map the same colors etc. onto the same numbers/letters/sounds/etc.

    I sort of found a priming effect somewhat similar to what you’re proposing in my PhD research. I was looking at all the influences that resulted in people pronouncing ‘MeFi/Mefi’ one way or another. I found a positive correlation between people who often wrote ‘MeFi’ with a capital ‘F’ and preferring the ‘Me fie’ pronunciation. Those who often wrote ‘Mefi’ with lowercase ‘f’ significantly preferred a ‘meffy’ or ‘meffai’ pronunciation. That is, people who recognised the capital ‘F’ and wrote it as such tended to process the word ‘MeFi’ as two distinct parts with a syllable boundary in the middle, resulting in a long, stressed ‘e’ sound. Lowercase ‘meffy/meffai’ people didn’t seem to “see” or process a boundary and their pronunciation choice reflected that.

    Anyways, it’s all a long rambling way of saying that there is a whole lotta crap that goes on when we visually process language, from what the letters look like, how they’re presented, what they mean, and what we expect to come next. And a whole boatload more.

    Hope you don’t mind this long comment. I always have time and interest in going on about this stuff!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>