Sounds like you should go to ATypI-Iceland! :-)

One thing you might factor in is divergence from the "P".
So maybe drop the bottom half-serif only? You'd then
lower the bowl to below center.

hhp

This is a problem... and a pain... bah!

In bolder weights it is sometimes not possible to make a reasonable Thorn without stripping off the serifs (as much as I hate to). And often there is no space to make a bowl without making the whole character look like some kind of a big ink blob but that is what one would always try to first.

The bowl does not have to be the size of the P bowl and often I rather watch the bowl of the R.

And yes, please visit us at the ATypI in September.

I've only had this circumstance once and didn't like the "missing serif" scenario, so I cheated by making the right side of the serifs thinner.

I figured that there are other strokes in the font that are thinner than the normal serif (here, in Eth), and other thin white spaces (e.g. between O and acute), to provide resonance—and as long as the glyph reads OK at a glance, no worries.

Thorn used to be with a descending stem when it was handwritten, so it's not so crazy idea. But this is hardly to be seen anymore.

It’s Ernestine, no? Just expanded character set.

> ... the Thorn can be allowed to look a bit smallish (in this case, narrower) too

Ah! Good thought. Thanks.
Seems like these serifs will get another chance! :-)

Jasper, Frode: Yes, Ernestine. It's being completed and finalized, so yes, it will be out before the world ends. :-)

it will be out before the world ends. :-)

And there was much rejoicing! :-)

I wonder if you couldn't build on Nick's approach by allowing the joins to become less horizontal and more diagonal ( like the join on the n) as the font becomes bolder which would allow you to have a bigger counter. Or you could angle only the outer portion of the bowl.

If you choose an approach like these looking at the joins in your very dense glyphs like th & 8 and anywhere else you may have made an optical adjustment would give you a kind of baseline to compare against.

Interesting thoughts Eben – I hadn't thought of trying such funny things with straightly orthogonal joins.

> one Gunnlaugur Briem has written a web scroll on the making of the thorn

He's quite brief about the cap Thorn though, and does not address this serif question. Unless I'm looking in the wrong place…

People do all kinds of wacky things to letters as they get bolder and bolder.... Still, the best thing is probably to find a group of users of the glyph and to show them some options ins seal setting and see which ones work best for them.

I suspect that the value of the serifs in making a rapid and correct ID of the letter is limited and that the clarity of the bowl and it's placement at the midpoint is the main thing. I can imagine loosing the serifs starting a a certain size to preserve what might be most essential.

Nice question and a nice thread!

I am curious to see if anyone else comes up with a still better solution.

Oops, looks like you're right. Thanks for the correction.

For what it’s worth, Jens’s example is also how Cyrus Highsmith chose to solve it in Salvo Serif Black:

 
I can’t say whether this succeeds for native Icelanders or not.

I have asked some Icelanders to come comment. Crossing my fingers....

Reminds me of what Colin posted to Twitter, from Photolettering: Þj Pj

I'm not an Icelander, but for my feeling it is rather too close to a P – especially since in my case I'm not doing a display font where that probably wouldn't matter so much.