Turns out people don't know what g looks like

Ori Ben-Dor

Why do we even bother?

https://hub.jhu.edu/2018/04/03/two-versions-of-letter-g-brain-science/

Chris Lozos

Makes perfect sense to me.  Perhaps if they were asked to "draw" this g instead of write it, they have changed the outcome,

Johannes Neumeier

Interesting. I'd like to see that study continued by showing participants four words each with a different g variation. I'd be curious if the detection rate increases.

Also, it shows how far removed from handwriting the looped g is to the writers of this day and age.

Mark Simonson

Normal people don't look at letters or think much about their shapes, as long as they can read it. I think in the case of the two-story g, the presence of the double bowl is the important part for recognizing it, the other details don't really matter. I know of at least one typeface where it consists only of the two bowls (I don't recall its name) and it works.

Christian Thalmann

Some typefaces vary the topology of /g rather liberally, and there's pretty much a continuum of shapes between the mono- and binocular paradigms. The Koch /g and Lato's italic /g come to mind.

As for «just two bowls», Quinoa's variants come close:

Kent Lew

I wonder whether if you showed a group of people four cropped photos of noses only and asked them to pick out the one that belonged to their best friend, you might get similar results.

And I wonder whether they’d do any better if they’d been taught how to sculpt portraits of their friends out of clay when they were growing up.

Ori Ben-Dor

I haven't noticed Lato italic's /g before. Its ear faces the right direction, but the loop is  mirrored. Interesting. Personally I don't like it, but it doesn't look completely wrong, probably because it's not far from a single story g.

It's also worth mentioning Benguiat's /g, which reconstructs the single story g from non-standard strokes.

Jens Kutilek

The same effect can be observed when you ask people to draw well-known logos from memory: https://www.signs.com/branded-in-memory/

John Hudson

Or common objects such as bicycles.

Jasper de Waard

John Hudson said:

Or common objects such as bicycles.

Indeed. Though some of these look pretty cool. https://www.wired.com/2016/04/can-draw-bikes-memory-definitely-cant/

With regards to the g, I wonder how odd g's influence legibility. 

Craig Eliason

Kent Lew

Wow, Craig, that vintage sure has some legs. ;-)

Stephen Coles

Mark Simonson said:

I know of at least one typeface where it consists only of the two bowls (I don't recall its name) and it works.

Maybe you’re thinking of Ladislav?

Chris Lozos

Here is my Dez Petranian:

Mark Simonson

I think it was Vero:


Not exactly a classic. Don't think I've ever seen it anywhere but in a type catalog. I suspect there may have been many more weights and styles. (And look: Tall instead of Condensed!)

AbiRasheed

@Craig Eliason Sorry man but holy smokes that g looks horrible specially for that sort of raised UV printing. 

Nick Shinn

Teaching type design to graphic design students, in University, I found that many of them instinctively placed the crossbar of /f low, as in Antique Olive.

Nick Shinn

Always impressed with Poor Richard’s /g (and other characters!)

Simon Cozens

It wasn’t until trying to design a /k that I realised I had *absolutely no idea* how they worked.

Adam Jagosz

I wouldn't say Lato's g is flipped, the stroke connecting the bowls is curved the same direction as in regular g's. I'd say it was rather translated to the right by a decent distance.

Wolf Böse

same problem exists with bicycles
http://twistedsifter.com/2016/04/artist-asks-people-to-draw-bicycle-from-memory-and-renders-results/