{.}

Apparently (via Twitter) James Todd is making a Didot where the text cut's Italic is ~15 degrees while the display's is ~18 degrees. Based on this:
https://twitter.com/JamesToddDesign/status/341331468628422658/photo/1
But note that this is the opposite of G&H's Mediaeval-Kursiv!

We need to think about this.

hhp

Whether one varies the italic angle at all, or in which manner one varies it, must depend on the individual design and what your intentions are, especially at the smaller sizes: is one trying to emphasise the 'italicness' of the letters by slanting them more, or is one trying to improve the legitibility of the letters by making them more open and less compressed and hence slanting them less? One can easily conceive, surely, that a sloped roman type of italic would benefit from more slanting at smaller sizes, while a highly cursive design might benefit from less slanting at smaller sizes. At display sizes, everything is pretty much a matter of style and aesthetics.

It is indeed more complex than it seems. It involves what I call the "drama axis", plus the extent to which an Italic should be stand-alone versus subordinate.

BTW, thanks to Lars Schwarz* here's a scan of that G&H face (with the captions aligned left for visibility):

* http://www.flickr.com/photos/bitrocker/8887501725/in/pool-type_specimen

But is 20-28 point a sensical cut-off for text-versus-display?

hhp

What if this doesn't come from a design, but a technical background? Maybe the 28pt wasn't properly fitting on the body?

There could be any number of reasons for the difference in italic angle at this size, to do with technical capabilities as Lars suggests, or even involvement of particular customer requirements.

Lars, do you mean that the larger size(s) slant less because the (metal) kerns might break? Except smaller sizes are actually more prone to breakage; plus smaller sizes need to be more "automatable".

BTW, check out this puppy:
http://typophile.com/node/88438#comment-512548

John: You don't mean that a customer will specify an Italic angle variance, do you?

hhp

Hrant: You don't mean that a customer will specify an Italic angle variance, do you?

Not exactly. Consider that some sizes of a type might only exist because a particular customer requested them and paid for their manufacture. So while it is unlikely to be a case of a customer specifying a variance in the italic angle, it could be a case of a particular customer having requested an italic style of a particular slant for the size they needed, and then that being incorporated into the family despite the italic angle variance. These things happen.

With regard to technical constraints, consider the possibility that the casting equipment might have a horizontal aperture limit, meaning that as italic type gets taller it needs to be steeper in order for the widest letters to fit within the manufacturing width limit.

I don't know if either of these possibilities is the case: one would need access to the company archives in order to determine the reason for the variance.

a horizontal aperture limit

Ah.

as italic type gets taller it needs to be steeper

Wait, do you mean less steep?

hhp

Steep = closer to vertical

I knew I'd come across this recently. This comes from The Making of Garamond Premier by Robert Slimbach:

...the lowercase letters of Granjon's Petit Texte size are significantly wider and more upright than his more stylish text size italics. By opening the lowercase counters and decreasing the italic angle, Granjon was able to improve legibility significantly.