Upper store of the lc a

Vasil Stanev

Hi beautiful people, 
since some time I have begun to notice that the two-story form of /a in some fonts has a hook that is, too my eye, imbalanced compared to the counter. I see it pushed to the left so much that I have begun to wonder if I am not missing something. We are talking not only recent fonts, but classics like Garamond.

To my knowledge upper parts should be almost always smaller then the bottom ones, like in an /eight, and an /a should stand balanced on its own as well as in text. Also, an extended hook should need something to balance on the bottom right hand side, and that's often not the case.

Am I missing some important peace of information or is it just people not knowing the rules? Please elaborate.

Chris Lozos

There is no absolute,  The upper and lower need to balance each other as well as fit with other glyphs like 's" and "e"

Mark Simonson

It's not the same principle as with the /eight or /S. Those are symmetrical characters conceptually. The top part is smaller to make them more optically symmetrical. The /a is not symmetrical in that way. The size of the bowl on the /a is often more related to the size of the eye of the /e, and this is especially the case with Garamond.

Thomas Phinney

Agree with Chris and especially Mark. There are two parts to this question of balance with the “a”: the size of the upper aperture compared to the counter, and the left-right balancing/positioning of the arch.

I am glad to share an insight I got from one of my mentors regarding left/right balance in lowercase.

Back ~ 16 years ago, I was working on a typeface (that eventually became Hypatia Sans) and showing my work in progress to Robert Slimbach. He pointed out that I had clearly worked very very hard to make my S perfectly balanced with no lean whatsoever... and now I should tweak it so that it leaned forward just the tiniest bit. Deliberately unbalance that letter just a little. The typeface as a whole was looking too static, and needed a little life breathed back into it.

It is important to realize that in many or most Latin (and Greek and Cyrillic) typefaces, some letters do not have to be perfectly balanced, and that is not only acceptable, but can be actually desirable! This is particularly true of the S/s optionally leaning forward just a tiny bit (but not left).

Similarly, the “a” does not need to be completely static and perfectly balanced. In some typefaces, even ones designed by masters, it seems to lean a little, left or more commonly right by not having the left terminal balance the right side of the arch. That is not usually an error or poor crafting on the part of the designer, but rather a design decision.

Look at Hermann Zapf’s typefaces (e.g. https://www.myfonts.com/person/Hermann_Zapf/)  Many have a very balanced “a” (Palatino, Aldus, Melior, URW Antiqua, etc.). But check out Zapf Renaissance Antiqua, ITC Zapf International, Marconi, and maybe Comenius and Optima. Not balanced at all.

Ori Ben-Dor

Another example of an "unbalanced" lower case a that works: Gill Sans. Compare it to Mallory, for instance. They take completely opposite directions, and yet both work just fine.

A similar difference between serifs: Adobe Text vs. Palatino.

Perhaps the most important observation here, in my opinion, is this: Concerning letter shapes, "right" and "wrong" isn't an intrinsic matter, or at least not just. It depends on traditions, on how any new letter shape refers to and echoes existing ones. If there's a tradition of some sort of "unbalanced" lower case a's, which is clearly the case here, then that becomes "right".

To illustrate that idea, referring to Mark's comment, note that while S, s and 8 are conceptually symmetrical, still noticeably smaller upper half, which breaks the visual symmetry, has become traditional and therefore legitimate.

Austin Stahl

When I was first learning to distinguish typefaces by eye, I realized that Caslon's dead giveaway is the /a that looks like it's lunging leftward. I always like when I see a design that seems to reference it.

Quirks like that are great when they're executed confidently, I think.

LeMo aka PatternMan aka Frank E Blokland

The size of the bowl on the /a is often more related to the size of the eye of the /e […]

Which is not surprising if one looks at the underlying construction coming forth from the movements with the broad nib:



The /ae diphtong shows the relation nicely.

Nick Shinn

There’s not much point in considering a letter shape in isolation.
Letters are social creatures.

LeMo aka PatternMan aka Frank E Blokland

Not surprisingly, because of the organic morphological relationship between the /a and the /e, in Garamont’s Gros Canon Romain (which appeared for the first time in1555) these letters share the same width (together with the /c). Sharing character widths is, of course, also the case for other letters, such as n and o, as can be explored with LeMo. The justified (for fixed mould-registers)  matrices on the photos are from the collection of the Museum Plantin-Moretus.



BTW, one can clearly see that the /a leans a bit backwards. This is for the simple reason that otherwise it will optically incline forward. The same (actually reversed) is the case for the /t.