Capital letters add a lot to text in various ways, for names, brands, emphasis (although capitalisation for emphasis is frowned upon)

Why then, are there no upper case numbers?


5 Answers 5


They do. The thing is, you probably don't realise, because upper case numbers have been all you've been using or seeing.

There is a distinction between 'default' numbers and 'oldstyle' numbers. The default numbers we all know are the actual capitals, with the 'oldstyle' numbers (sometimes incorrectly called 'proportional numbers') are lowercase.

Fonts tend to default to one style or another. Most font files allow you to change default numbers into oldstyle ones by turning them into 'small caps', but you can also select them from the Glyphs palette (Shift+Alt+F11 in InDesign).

Don't confuse 'default' vs. 'oldstyle' with 'tabular' vs. 'proportional', that's a distinction in how the numbers are lined out horizontally.

enter image description here

This article outlines the difference between the kinds of numbers, and also gives some tips on how to achieve the different looks in InDesign.

  • 5
    This is a good post. May I please offer an improvement to make it even better? Say “lining” figures not default figures. Calling figures default ones doesn’t say whether they’re text or lining, proportional or tabular—and a font’s default figures may be any of those four. (Bringhurst usually calls oldstyle figures “text figures”, so I’ll follow his conventions here.) While a typewriter’s figures are always tabular lining figures, the typesetter has 4 possible choices: ❶ proportional text, ❷ tabular text, ❸ proportional lining, or ❹ tabular lining. i.sstatic.net/VUG5e.png
    – tchrist
    Commented Oct 6, 2018 at 17:12
  • 1
    The specimen sample is from EB Garamond. The authors there further write: “This font is primarily intended for texts. Thus default figures are oldstyle proportional so they fit best in a text setting. For many purposes though, tabular (all figures have the same width) and/or lining figures (they are about caps height) are preferred. They can be accessed using the features tnum and lnum.” Unlike EB Garamond’s proportional ones, Georgia uses tabular text (=oldstyle) figures as its default figures.
    – tchrist
    Commented Oct 6, 2018 at 17:16
  • Here’s another specimen sample showing all four combos in each of Minion Pro, Adobe Garamond Pro, Adobe Caslon Pro, and Myriad Pro. Not all fonts offer such niceties as four possible figure combos or actual small caps, but these four do so. Feel free to use the sample if you’d like, or make your own if you prefer.
    – tchrist
    Commented Oct 7, 2018 at 19:06

While upper case numbers do exist, as is shown in vincents answer, they did not originally exist at all. Remember, our numbers are copied from the Muslim scientists who wrote in Arabic.*

Arabic is unicase. That is, all letters are same case, so the notion of big and small numbers is a later development. Since the original system had no case, neither did the adopted system.

It might be worth noting that the roman numeral system also did not have small letters or numbers. This was a later development to make writing easier. Uppercase letters were designed for stuff like engraving. Since the numbers were copied from a system that was meant to be written with a pen there was no pressure for change. You can see this from the fact that upper and lowercase numbers are the same with just different height characteristics.

* Some people want to point out that the numbers came from India. True the Arabs borrowed the idea form the Persians who borrowed them from India. Most notably the Zero was invented in India which makes the system work. Despite this the modern numerals were a rehash on this idea and the modern form was developed in northern Africa and are distinct from the eastern numerals. Since Europeans copied them, as is, without modification it can be said they swiped the design from the western Muslims.

  • 7
    The Indians invented our numerals. While you are technically correct, this answer gives the impression that Arabs invented the Hindu-Arabic numeral system.
    – user46532
    Commented Feb 1, 2016 at 5:56
  • 2
    @jasonszhao Added the comment so its stays for posterity.
    – joojaa
    Commented Feb 1, 2016 at 7:29

What is case?

The discussion both in this question and in the one it inspired on ELU seems to conflate two distinct meanings of ‘uppercase’ and ‘lowercase’:

  1. Based purely on shape and size, originating in whether a glyph was originally usually stored in the typographer’s upper or lower case (= drawer).

  2. Based on functionality, describing what upper- and lowercase letters are used for in modern-day English.

When we think of uppercase letters, we no doubt think of both of these at the same time: A is not only recognisable as something that is distinct in its form and shape from a, it is also distinct in that it is used at the beginning of sentences, in acronyms, to start names, and (in good taste or not) for emphasis and headers.

Numbers are a somewhat different matter, though. If we define case by (1) alone, then it makes perfect sense, as in Vincent’s answer, to say that lining numbers are ‘uppercase’, while text numbers are ‘lowercase’—the picture linked to by curiousdannii even shows that the different styles of letters were frequently kept in the upper and lower cases in typographers’ kits.

But if we define case in terms of function, it is abundantly clear that lining and text numbers do not correspond to upper- and lowercase letters. The use of one or the other type of numbers is dependent on entirely different things, and they are never mixed the way upper- and lowercase letters are.

Unlike Vincent, I consider the functional aspect very central in the modern-day definition of what ‘case’ is, and I consider that lining and text numbers are not ‘upper-/lowercase numbers’, at least not as we would normally understand that. In other words, as presupposed by the question, that there are no upper- and lowercase numbers for the present purposes.


Why is case?

The original shapes of the letters of the alphabet (all the way from Egyptian hieroglyphs through Phoenician, Greek, and eventually Latin) were what correspond more or less to our current uppercase letters. They are thus the ‘default’ letters, historically speaking. They were made originally for scratching, etching, hewing, engraving, and otherwise writing on very solid materials: stone, wood, clay and wax tablets, etc. The fact that writing in these non-cursive letters is quite slow didn’t matter much, because writing in the materials available was much slower in itself. At this point, the alphabet was unicase.

Cursive forms of the letters arose (some from variant forms of the previous unicase glyphs, some from inventions, some in various other more or less complex ways) when more sophisticated ways of writing became available and writing in big, heavy block-glyphs became too time-consuming. In other words, when the bottleneck had moved from being the material used to being the glyphs used.

Initially, these cursive letters were simply variants used when quick writing was desired; for more ‘formal’ writing (engraving on statues, etc.), the older non-cursive forms were still used. The alphabet was still basically unicase, since the two different ways of writing were not mixed—they were basically like two different fonts.

It wasn’t until much later (I’m not sure exactly when, but some time between the 7th and 12th centuries, approximately) that the two letter styles really began to be mixed together in a system more or less comparable to our current upper- and lowercase one.


Why no numeric case?

This leads us to the actual question: why did no cursive (= lowercase) numerals begin to emerge along with the lowercase letters?

No authoritative answer is likely to emerge on that count, I’m afraid—the full reasons are lost in history. I would venture the following hypothesis, however:

  • When cursive letter forms began to emerge and be used as an alternative writing style (in Latin, some time during the first century BC), the Arabic numbers—themselves borrowed from Brahmic scripts in India—had not yet been borrowed; they used Roman numerals instead, which were just letters and thus had cursive variants

  • At some point, Arabic numerals were borrowed and started supplanting Roman numerals

  • While the cursive and uncial styles were in concurrent but complementary use, cursive and uncial styles of the Arabic numerals developed that each fit their own letter style

  • When eventually the modern case system emerged from these two variant styles, the function that was attributed to both was not really felt to be relevant for numbers (write a name with a capital number? Begin a sentence with a capital number?), so they continued to be just variants of each other, to be used where their shape fit best



Addendum: other scripts

Uppercase letters have one more function that has not been mentioned: clarity and disambiguation. For example, when filling in a form, it is common to see “Please write in block letters only” or something to that effect, to make sure that what you write is clear and legible and cannot be mistaken for something else.

To some extent, we do have ‘uppercase’ or ‘block’ numerals for this function: I (one) gets a serif at the top (1) and a horizontal stroke at the bottom (1̱); 7 gets a middle stroke (7̵) to disambiguate it from 1; 0 (zero) gets a slash through it (0̸) or a dot inside it (can’t figure out how to display that here) to disambiguate it from the letter o.

Some scripts that do not distinguish case do actually have a distinction in numerals that mirrors this. In Chinese, for example, all numeric characters up to 10,000 (i.e., all the characters you need to write up to 99,999,999) have two forms: a ‘short’ form and a ‘long’ form. The long form is used when absolute clarity and nonambiguity is necessary, e.g. on cheques (where they also help prevent fraud by adding an extra stroke to make a 1 a 10), here given in simplified characters as short form / long form (the bullet at the beginning is zero):

〇 / 零

  1. 一 / 壹
  2. 二 / 贰
  3. 三 / 叁
  4. 四 / 肆
  5. 五 / 伍
  6. 六 / 陆
  7. 七 / 柒
  8. 八 / 捌
  9. 九 / 玖
  10. 十 / 拾
  1. 百 / 佰
  1. 千 / 仟
  1. 万 / 萬

Interestingly, the difference between these two styles is known as 小写 xiǎoxiě ‘little writing’ and 大写 dàxiě ‘big writing’, which is also the terms used to translate the Latin alphabet–based notions of ‘lowercase’ and ‘uppercase’.

  • See John Crossley’s “Old-fashioned versus newfangled: Reading and writing numbers, 1200–1500” in Studies in Medieval and Renaissance History, Third Series, Volume 10 (2013).
    – tchrist
    Commented Oct 6, 2018 at 18:24
  • 2
    It turns out that “number case” is a ᴛᴇʀᴍ ᴏꜰ ᴀʀᴛ in TrueType/OpenType fonts, the very one suggested by Vincent’s answer above. Quoting from Adobe’s TrueType Reference Manual: “Number Case is independent of Letter Case. Lower Case Numbers (also called "traditional" or "old style") are digits which may descend below the baseline, as opposed to Upper Case Numbers (also called "lining"), which do not descend below the baseline.”
    – tchrist
    Commented Oct 27, 2018 at 17:49
  • "For example, when filling in a form, it is common to see “Please write in block letters only” or something to that effect, to make sure that what you write is clear and legible and cannot be mistaken for something else." Which is ironic, given that all-caps is actually less legible and harder to read than sentence-case text (due to the lack of ascenders and descenders to help the eye along).
    – Vikki
    Commented Dec 28, 2021 at 4:50
  • @Vikki In typed text, yes. I was talking about handwritten text here, where the benefits of a- and descenders in sentence case are vastly outweighed by the text being joined up, written at higher speed and with more idiosyncrasy, making block letters far more legible in most cases. Commented Dec 28, 2021 at 8:21
  • @JanusBahsJacquet: That would've been accurate up to about fifty years ago, but nowadays hardly anyone uses joined-up handwriting except in signatures (which don't need to be legible, just distinct).
    – Vikki
    Commented Dec 28, 2021 at 8:29

Capital letters exist as our written and printed language has decided they should.

The rules for usage of capital letters typically is for starting sentences and proper nouns.

The rules simply don't apply to numerals. Hence, no need for there to be 'upper case' numbers.

Your example of using ALL CAPS TO SHOW EMPHASIS is actually not an ideal way to show emphasis. Most typographers would suggest you use italics, boldface, smallcaps, or even color instead. And most typefaces that include those variations tend to also include numerals in those variations.


The core point is that numbers is a different semiotic system than letters.

Am i completely out there in thinking that the point of numbers is math, which is a "language" all of its own. An abstraction for trade, money, calendar-time, distance; crossing languages in the ancient world etc. (I do however love the oldstyle numbers @vincent demonstrates).

Writing letters on stone in capitals are a lot easier to calculate sizes - i.e. romans. Writing by hand on paper needs to be faster, hence cursive; and the idea of differentiating with capital vs. "handwritten" to make reading the images of words easier (exactly when that happened I do not know).

There are examples of no uppercases and no punctuation in manuscripts. Math, on the other hand, does not have this issue. We ditched the roman numerals, as the arabic have 0, and roman numerals do not (and fractions gets interesting with roman numerals). And I suppose you do NOT want your trade calculations, your debt to be ambiguous; the sleigh of hand of your accountant to be up for discussion.


Not the answer you're looking for? Browse other questions tagged or ask your own question.