Typefaces (Fonts) for People with Reading Difficulties

You might think that a passage of text that can be read easily by someone who has difficulty reading passages of text, would point to a clear-to-read typeface. Maybe it will, but the writings on it can be a bit unconvincing.

For example see, Typefaces for Dyslexia, at www.dyslexic.com:80/fonts which makes the rather surprising assertion that the serifs found on traditional letter forms ‘tend to obscure the shapes of letters’. Pretty revolutionary, considering how centuries of practice have tended to indicate the opposite (see About Legibility by Adrian Frutiger).

May I recommend that you read The Science of Word Recognition from Microsoft Typography, by Kevin Larson. Though it begins with a rather amusing-sounding sentence: ‘Evidence from the last 20 years of work in cognitive psychology indicate[s] that we use the letters within a word to recognise a word.’ (My copy-edit in [ ] brackets). Gosh! That’s profound research isn’t it? But actually as you read on you begin to see that the sentence has a serious meaning, it’s just (inadvertently I’d guess) put in a way that sounds funny.

Kevin Larson’s paper talks about experiments in word recognition in terms of overall word shape and letter identification. It is very English-language oriented, and I would have welcomed some indication of experiments undertaken using subjects who did not know the meaning of the words they were being presented with, to see how much difference there would be in the findings, but all-in-all it does seem, from the experiments that have been done, that fluency of reading is much greater, when the subject is familiar with the content. Note that this doesn’t mean that they understand the words necessarily, more that they are familiar with the contextual content – you really need to read Kevin Larson’s paper for an explanation of what I mean here.

The implications for readability by people with a reading disadvantage are not touched on in Kevin Larson’s paper, but much may be inferred from it.

But then you read on and see that they say that serifs are ‘found in traditional print fonts such as Georgia or Times New Roman’, and you think, oh dear, those aren’t traditional print fonts, but Microsoft fonts. It’s the old story, someone has got hold of their computer and they think it’s really shiny.

You will most probably find it easier to read a classic-style book set in a serif typeface than in a sans-serif one. This is especially so when there are a fair-sized number of words on the page, such as in, for example, a novel. Try it for yourself and see (if you can find a novel-length book set in a sans-serif font that is, which will not be easy). The reason may be simply because that’s the format you are used to. The research that has been undertaken consistently indicates that there is no difference in readability between serif and sans-serif fonts, see www.alexpoole.info /academic/ literaturereview.html but whether any of the research has included novel-length books is another matter. It appears not to have done.

Whether that observation, that a serif typeface suits a novel best, applies when a dyslexic person reads a novel, I couldn’t say. Maybe a person with dyslexia requires a special print run, with novels set in a sans-serif typeface that the majority of people would find it more difficult to read than the usual format, but I doubt it – I suspect that some of the guidelines are ill thought-out. All the indications are that there is no optimum typeface for readability, it very much depends on what you are reading.

Longish passages of text in regular patterns certainly can look jiggety in a sans-serif foint as opposed to a serif one, see my Arial Creates Optical Illusions page and my Font Readability Experiment page. Notice that these experiments (that anyone can do at home) would tend to point in the exact opposite direction to many of the supposedly expert guidelines, indicating that it is the serifed fonts that are easier to resolve, but again, and again, I must emphasise that it very much depends on what you are reading.

Read Regular is a typeface designed by Natascha Frensch at the Royal College of Art in London, specifically for dyslexic readers and has gained some level of notoriety.  (www.readregular.com). The claims for the Read Regular font as being better-read by people with dyslexia do not appear to be backed up by any independent evidence. Natascha Frensch has tried to make each letter a different shape from any other, so for example, d is different from mirror-image b.  But it’s not the only typeface to do that by any means. It appears that you cannot buy Read Regular but have to have the typesetting done in the studio of Natascha Frensch. Clever, eh?

In fact that feature much shouted about in Read Regular (and about some other fonts touted by www.dyslexic.com:80/fonts), that the shape of a letter b is not a mirror image d, applies to lots and lots of fonts. If you look at Myfonts.com and type ‘sans’ into the search box – I just did this and I can see from the results that the sans-serif fonts with non-mirror lower case b and d include: Alinea sans, Vista sans, Priori sans, Fedra sans, Benton sans, Ela sans, Relato sans, Freight sans, LTC Goudy sans, Placebo sans, California sans; etc etc etc, hundreds and hundreds of them, most of which you’ll never have heard of and the reason you haven’t is because they have not had a weight of marketing behind them. (And this is one of the reasons that I can confidently say that www.dyslexic.com:80/fonts is full of nonsense).

And if you look at serif fonts, they nearly all have a ‘d’ that isn’t a mirror of ‘b’, including Times New Roman: b d.

I’ve often heard it said that people with reading difficulties or disabilities find the letter forms that more closely relate to handwriting easier to deal with. This is especially significant in the lower case letters a and g. In a serif font like Times New Roman, say, a and g typically look like this:    a g. Or at least they do in their regular or Roman form; in the italic form they look like this:   a g, and there are some serif fonts, for example Bookman and Microsoft’s Georgia, where the Roman forms of a and g look like this: a g, and the italic forms look like this: a g. Notice that in the regular or Roman form of Bookman and Georgia the lower case a has three tiers. In the italic form, by contrast, the a is like an o with a line to the right, and the g has a tail rather than a loop.

(Note than in the above I’ve used text rather than graphics, so if you don’t have Times New Roman or Times, or Bookman or Georgia on your machine, you may just have to imagine it and believe me).

Sans-serif fonts vary. Some have the three-tier a (as do Microsoft’s Arial and Verdana) and some have a calligraphic-style (sometimes called infant-style)   a. Sans-serif fonts typically have a non-loop, tailed g, but not all of them do. Again a random selection from Myfonts.com, the following fonts have a looped g like you typically find in Roman serifed fonts: Priori sans, Leitura sans, Relato sans, Freight sans, LTC Goudy sans, California sans, Alinea sans, etc etc etc. Lots and lots of them. If you keep your version of Windows XP up to date or you use Vista, then you probably have a font called Calibri and another called Candara, which are sans-serif fonts with a three-tier g.

Read Regular is one of those sans-serif with a calligraphic-style lower case a that is, as with Comic Sans, very similar to lower case o, (a and o in Comic Sans) and is one of the fonts that makes pretty much zero distiction between an upper case I (I) and a lower case l (L). That may be an advantage for dyslexic readers, if the authorities say so then it must be. Seems a bit unlikely, but there you go.

Does it strike you as you read this, that websites such as dyslexic.com, together with those people who maintain that this or that font style is better for people with reading difficulties, still have a bit of mugging up to do?

Having said that, though, if you were to lay out some text to be read by a person who has a reading difficulty, and you used a font with a looped    g, you can be sure that some expert educator or another will tell you that this is not correct, that the target audience will not be able to read it easily. The research from the University of Reading, Typography for Children, is interesting in this respect, it found that there is no evidence that says that either serif or sans serif typefaces are intrinsically more legible, but teacher opinion, generally, favours sans-serif typefaces because of the ‘simplicity of the letter shapes’, by which they presumably mean the similarity of the letter shapes to those that the teacher uses when presenting handwritten words to their pupils. (For as you can quickly establish from a look at, say, Myfonts.com, there is no specific letter shape that is exclusive to a sans-serif font, except perhaps for the absence of serifs, which might be considered to make the letter shapes a bit more simple, possibly, sometimes.)

There may be something in the argument that letterforms that look rather similar to those that have been taught, will be better received by people who have a learning disability. I would imagine that this does not apply to someone with dyslexia, but for a person with learning disabilities, it may be a point to bear in mind. So far as I am aware, no real research has been done on this. And it begs the question, what typeface looks like the teacher’s handwriting, when you do not know what the teacher’s handwriting looked like?

But we’ll take a guess. Here are two passages, the first in a font called Architect Small Block, and the second in Comic Sans. These are graphic images so the fonts will be accurate and not dependent on what is on your computer (for more details on the rendering of fonts on a web page, see my page on Fonts on the Web).

Now while Comic Sans is much beloved by many teachers and self-promoted readability experts, in fact you’ll see from the above passages that relative to the font to its left it has quite a number of fancy letter features, the other passage is more like what you imagine the diligent teacher would have written on the board. Probably, the belief you hear sometimes about Comic Sans being highly readable is really that the teachers, well, they just like it.

And now, in colour. Evidence seems to indicate that many people, not just those with dyslexia but also others who find reading to be a harsh experience, find a passage of text easier to read through a coloured filter or coloured spectacles. This is especially associated with a perceptual problem known as Irlen Syndrome, see www.irlen.com.

Those who treat reading problems using coloured filters or glasses consistently report encouraging results, though it seems that no one yet knows why it should work. One effect of the filter will be to reduce the relative brightness of text against background, and it may be interesting to know whether text in a shade of grey, as opposed to text in black, on white paper has a similar effect, or whether black text printed on grey paper, or even on coloured paper, is equivalent. So far I can find no indication that anyone has asked the question. The assessment for the coloured filters is sometimes done by adjusting the background colour on the computer screen, but looking at text through a filter will adjust the foreground colour too, which the computer screen adjustment will not, or not unless that adjustment is done too, and again this seems to point to a relative brightness issue, more than one that is specifically related to colour, though obviously more research is still to be done.

As anyone knows who has sat and read a newspaper under a tungsten lamp – ie just about everyone – adjusting the colour of the ambient light is something we soon get used to and don’t perceive as especially different from daylight conditions. So this reading through coloured filters is awash with questions. Obviously if it helps, then that’s good, but getting an explanation seems to be elusive.

I should say here something about colourblindness, which is a form of disability I suppose. http://colorfilter.wickline.org is a mighty clever and useful tool that lets you look at your website – or someone else’s – through a colourblindness filter. Very useful for testing one’s website indeed. You may be surprised, or not I hope, to find that the majority of web pages that are hard to read with a filter turned on, are also hard to read with it turned off. This is not to deny that those with, say red-green blindness cannot distinguish between red and green, rather it is that red on green at too close a brightness level to each other, while perhaps impossible to distinguish by someone with a colour perception deficiency, are also pretty hard to read by anyone at all. The following two pairs of blocks are in red on green, and I would make bold enough to say that someone with red-green colourblindness can read one of each pair easily enough, but will have a lot of difficulty with the other. It’s to do with relative brightness. In the first pair of blocks, the red is in fact identical in both the left and right samples, even though it probably doesn’t look the same.

Read me well, sister!

Read me well, brother!

And here’s another example, where the green stays the same in the two samples but the red changes. The relative brightness has the greatest effect on readability, for as you can see, where the relative brightness of the colours is close, it’s not easy to read, even if your colour vision is 100%. Obviously more difficult if it isn’t, though with the rightmost examples, everyone can read it OK, yes?

Read me well, brother!

Read me well, sister!

You can see more about type in colour on my text colour readability page.