Ben Haskell wrote:
> I can adjust my immersive reading speed from "engrossed
> sci-fi novel" speed (slower) to "e-mail screening (do I
> really want to read this)" speed (faster).
Sure. And by doing that you're rationing your reading comfort,
so to speak. But we can't account for that. And that's why
you have to make your test subjects read a *lot* of text
when timing them.
> I could get good "readability" scores from nearly
> any font, just by concentrating and reading faster
> than normal.
But after dozens of pages you'd be exhausted,
if the font had low readability.
> So I don't think that "speed alone" is
> sufficient to define "readability".
Not momentary speed - speed over dozens of pages.
It's like the 100 meter dash versus the marathon.
A book is a marathon.
> We need a fatigue or endurance measure as well.
Yes, but you just incorporate it with "speed" when testing.
And when you think about it, that's as close to real-world
reading as you can get.
> I suppose that readability in terms of "fatigue-resistance" is
> ultimately reflected in reading speed...if enough pages are read.
> To get a handle on "fatigue", fonts could be tested
> under both optimal and high-stress conditions.
> Every font should be "stress tested" to its breakdown point!
"Rolf F. Rehe" wrote:
> Actually, under optimal reading conditions, fatigue does
> not occur for at least two hours, one study reports (I can
> cite the study if needed.).
Could you, please? It sounds interesting.
> Content, me thinks, plays a (perhaps major) role as well.
Very much so. The more longer and unfamiliar words a text
contains, the more fixations we have to perform, and the
more we get tired.
> One often overlooked element is the type/background relationship.
> Thin, hairline type is inherently of poor legibility in text type
There is, however, an interesting "side-theory" in
readability that says that letters should ideally
decompose into their parts, so that they contribute
to the bouma without being too strong an individual
presence. I have a hunch that Unger uses this, but
I can't be sure.
If this is true, then some degree of thinning of the
joins in letters might be beneficial. Thin straight
lines, however, are generally not good, I agree.
> What role does 'speed reading' play?
That's in interesting question. Maybe books intended
for speed reading need a certain kind of font?