Welcome to Typophile
Please Sign in.


Indices : Terminology : Figures

Arabic numeral glyphs. Often characterized as oldstyle, lining and tabular.

The first appearance of any numerals in a manuscript in Europe:
Codex Vigilanus, written by a Spanish monk in the year 978

The first printed books did not make use of Arabic numerals at all. None of Gutenberg’s printings contain any Arabic numerals. Ratdolt was one of the first to use printed Arabic numerals in his books – many of which were about mathematics and astronomy. His numerals left behind any traces of medieval forms and are immediately familiar to the modern eye. The zero is a perfect circle. Aldus’s zero was a perfect circle, as well as the zeros in Garamond’s typefaces.

Numerals with perfect circle zero by Erhard Ratdolt, 1482.

Numerals with perfect circle zero by Aldus / Griffo, 1502.

Francois Guyot was one of the first type designer (or was he the first at all?) to cut both a roman and an italic related to each other, evidently with the idea that both may be used together. He used exactly the same numerals for both the roman and the italic typeface (as one can see in the Folger type specimen of c. 1565). His zero is a perfect circle. (BTW, perfect circle actually means perfect circle - the counter in one of Granjon’s typefaces was drilled, as one can see in Fred Smeijers: ‘Counterpunch’, p 158.)

Ameet Tavernier was the first punchcutter who designed inclined numerals for an italic typeface, but he did not dare to touch the perfect circle shape of the zero, so the zero is upright and not inclined.

Numerals for an italic by Ameet Tavernier, 1554
The numerals are inclined, except perfect circle zero.

Approximately at the same time, Robert Granjon cut his first italic with inclined numerals, but he also slanted the zero, which therefore turned into an oval. It is very interesting to observe that Granjon, even if he abandoned the shape of the perfect circle, retained the other characteristic feature of the zero, its monolinear stroke thickness. He must have found it important to still keep the shape of the zero different, while trying to bring the design of the numerals closer to the design of the italic.

Numerals for an italic by Robert Granjon, c. 1554
The zero is also inclined, but still monolinear (without modulation in stroke thickness)

The introduction of numerals of equal width (one of the two features commonly attributed to the so-called ‘tabular figures’ as opposed to the so-called ‘old-style figures’) preceded the new design of numerals of uniform height. Already in 1764, Pierre-Simon Fournier le jeune wrote in his Manuel typographique about numerals of equal width:

‘The figures of the roman have no respective sets, but one uniform one. They are all cast to the thickness of an n-quadrate; that is to say that two of them put side by side make a square of the size of the body.’

The first numerals of equal height were cut by Richard Austin for the founder John Bell in 1788. It was the zero that suffered most from this treatment, since it was both squeezed into a far too narrow horizontal space and stretched to capital height at the same time.

To “oh” or Not to “oh”?
Lack of Small-Cap Height Lining Numerals

Additional Articles:
dot-font: Numbers Don’t Lie
FontFeed: Figuring It Out: OSF, LF, and TF Explained