And following up. If this “echo” of a different structure is repeated more than once through-out the alphabet, could the mixing of “vocabularies” work?

What about mixing styles within the same glyph? Here’s a sketch of an italic ‘n’ with rational inside and dynamic outside.

I would have to see more glyphs but I find the "n" to be pretty interesting. It seems to blend the two well.

Look for example at H&FJ's Mercury, which shows more than two different terminal styles. I certainly think there's nothing wrong with it, it is principally a matter of getting the contrast right and keeping consistency afloat. Getting the contrast right is more difficult when working on semi serifs (a successful example is Alcuin). As far as the mixing of styles goes, the task really boils down to mixing them consistently enough. Mixing styles will get in the way of achieving consistency, that's for sure.

I stumbled upon the same problem when working on my last typeface Meleo. What I learned from it is that if you're going to bring two different terminal styles together, you better make those styles clearly different.

I think you'll have trouble finding successful such designs because it's a very PoMo approach, and PoMo type has yet to mature. But there is zero reason this sort of thing can't work; historical faithfulness can be a crutch as much as it can be a shortcut to specific shape-solutions (and it's rarely much more than the latter). But of course breaking from precedent is more work, not least in terms of ensuring readability.

So even if something hasn't been done successfully in the past, it's still true that all you need is a reason to make it now. And lacking any evidence that it's actually functionally harmful, this reason could actually be as simple as: "I like the way it looks." And remember, unlike display fonts, text fonts work "under the radar", so much stranger stuff can make sense.

--

The visual discord between the arches and the "r" in your first example could actually work; but I think it's better if the "r" isn't all alone (which BTW is the main reason I prefer the open-bottom binocular "g" - it can accompany the "s"). Your second "n" is very nice, and promising. End of story. :-) However if you do want a precedent of how "wild" things can get, look at Fleischmann, especially his #65 (which people have yet to develop the cojones to properly revive).

Related:
http://www.typographer.org/2005_06_01_digests.html (see my comment, about halfway down)
http://typophile.com/node/13307 (with repeated apologies, but not for not covering it up)

hhp

Thanks. I’ve read Yves’ review multiple times before (it mentions Mercury, if I remember correctly). I also have a ragged old 1800-something edition of De Mensch typeset one of Fleischmann’s types. I’m not sure if I find it very relevant to this experiment though.

What I learned from it is that if you're going to bring two different terminal styles together, you better make those styles clearly different.
It’s not so much about the terminals, but that’s still relevant.

The visual discord between the arches and the "r" in your first example could actually work; but I think it's better if the "r" isn't all alone (which BTW is the main reason I prefer the open-bottom binocular "g" - it can accompany the "s").
Both branch structures are repeated multiple times through-out the alphabet. There’s a little bit of both ‘r’ and ‘n’ in ‘g’, for example, but not the spoon-fed kind. I suspect I might have too many ingredients though, with the Dwiggins-esque flips.

Btw: Will look closer at the Fleischmann structure.

What about mixing styles within the same glyph? Here’s a sketch of an italic ‘n’ with rational inside and dynamic outside.

This is somewhat related to what I'm doing with Millennial (in progress): building letters around pretty strictly symmetrical counters, but letting the outer contours be looser.

Craig, is it possible to have a closer look?

I’m ashamed of the rusty finishing of some of my drawings here. The concept is still up in the air, so polishing and fine-tuning comes later.

Frode, in case you haven't already seen this:
http://www.themicrofoundry.com/other/Fman65.gif

Craig, is this a new twist in Millennial? It used to be pretty conventional, at least in the Roman: http://www.typophile.com/node/83333
As you might suspect, I'm delighted to see it!

hhp

Craig, is it possible to have a closer look?

Check your email.

Craig, is this a new twist in Millennial?

Not terribly new, but yes, an idea that came up only in the course of working out the design. I can't remember if it came to me since my last contribution to that critique thread, or somewhere a ways into it, but either way I don't think we discussed the concept there.