First off:
Is it common to think that humanists are more legible than grotesks? Is it just something I made up in my head?
If not,
As I have recently discovered, if I take a grotesk and put a humanistic end on it, it doesn't lose any aperture size, if not decreases. Even when blurring, the grotesk /a holds together well.
Comments
Apocryphal wisdom is that open apertures and varied terminal styles better differentiate letters than close apertures and regular terminals. By this wisdom a humanist typeface like Lucida is on the very legible end of the spectrum and a neo-grotesk like San Francisco is on the other.
As to whether or not science backs this up, that’s another rabbit hole.
It still is a bit harder to read, but it's still as badly mutilated.
Supplemental image:
The problem is not with the two story a (it is distinct enough to ge legible in quite bad conditions) but with the more similar letters. Like 689, oec and so forth. Both groups are clearly distinct with humanist letters but can be a problem with grotesk shapes and bad conditions.
http://www.designmadeingermany.de/2014/2564/
Klick on the diagram to get the enlarged version.
That’s not how we see, nor how we think we see.
In the first place, the image that falls on the retina, during immersive reading, is not perceived with uniform sharpness, but is sharp for only about five characters, in a fixation in the fovea.
Secondly, the idea that characters which are outside the fovea are perceived as more blurry is incorrect. Certainly, their image on the retina there may be unfocused, but it is low resolution, more like a coarse halftone, perhaps something like this: (Pardon the grid screen, I don’t know the retinal pattern, it may well be stochastic.)
But this is not what we think we see. In processing visual images, the brain does a certain amount of guesswork, most notably in “cloning” over the blind spot from its adjacent area. Therefore it is reasonable to suppose that what readers think they see in areas of text outside the fovea is neither blurriness, nor low-res granularity, but a more refined proto-signifier, having both the quality of “lorem ipsum” pseudo-textual meaning (although in the language of the document), and generic glyphic “type-y-ness”, which is related to their experience with similar documents: in other words, the reader is anticipating a continuation of the same language, in the same typeface.
I propose the following “probability model” for Helvetica a-e-s decoding.
The three glyphs have been superimposed, each with a 35% opacity, the levels are tweaked, and there is only a slight blur. Remember, this is a representation of a stage in a mental process, because there is no homunculus inside us that can capture an image of what we perceive.
This image represents a momentary heuristic model that might be used in reading Helvetica, assuming the same probability for all these three letters as first perceived in words outside the fovea. Then, when we fixate on these letters in the fovea, we run a series of feedback loops involving letter details and possible words, refining disambiguation until closure on final signification occurs.
Here’s a scientific paper which discusses it:
http://www.sciencedirect.com/science/article/pii/S0042698906003932
To put it another way, I think your supposition is reasonable, but not for the reason you suppose.
That is not to say that there are no other variables relevant to legibility, of course. Also, "more legible" is only important in the situations where that degree of difference matters. In many situations and sizes, either will be legible enough.
The thing is, not everyone and all situations are equal. Why deliberately choose something that is going to fail more often?
On my little iPhone screen, when I tried to read numbers at small sizes in a moving car (such as the time at the top in small print, while the maps are in use), I cursed Apple for using Helvetica. I curse them only slightly less for San Francisco today.
Is this thinking true? (I don't have the typeface on my computer.)
Yes. Interstate is based on the old FHWA (Highway Gothic) fonts. Those fonts were designed for legibility at high speed, but not especially well. A better example of good highway signage would be James Montalbano’s ClearviewHwy, which is replacing old highway typefaces around the world.