14

I saw an episode of QI (Quite Interesting, a British 'quiz' show that just sort of presents trivia). I don't know the episode or when it was produced (I've searched for it on youtube but haven't found it yet).

On this episode one of the questions was "What symbol did the Ancient Romans use for 'one thousand'?". The answer, which everyone and myself thought, was "obviously 'M'".

But the show is all about contrarian facts. They said no, the 'M' was introduced in the Middle Ages and that the 'Ancient' Romans used some symbol that looked like '(I)' (with example photos).

Yet, that seems strange... I don't think I've ever seen anything other than 'M' used for the Roman digit for a thousand.

What is the real story?

Mitch
  • 727
  • 6
  • 14

1 Answers1

32

The shortest answer is, often they didn't use anything! Roman numerals weren't frequently used for numbers greater than a thousand; accountants doing math with big numbers would have alternate systems that made adding and subtracting easier, and year numbers past 1000 weren't a concern.

However, there were a few types of specialized notations for this. One of them is called the apostrophus system, where a straight line with curves on either side indicated larger numbers. ⅠↃ is 500, ⅭⅠↃ is 1,000, ⅠↃↃ is 5,000, ⅭⅭⅠↃↃ is 10,000, and so on. Here's an example of a year given in this notation, on a plaque on a church in the Netherlands.

a photo of ⅭⅠↃⅠↃCXXX

Sometimes these were combined into single glyphs, and thus you get Ⅾ, ↀ, ↁ, ↂ, and so on. The first of these was eventually re-analyzed as the letter D, which is how that became the symbol for 500. The rest generally fell out of use, and ↀ was replaced by M, the first letter in mille ("thousand"). (Though it's possible that the infinity symbol ∞ was, much later, adapted from ↀ.)

Another system, which became more popular for recording distances in Imperial times, was to put a bar over a number to multiply it by 1,000. In this system, a thousand would simply be I̅. This came about somewhat later, but still definitely qualifies as "ancient Roman".

Draconis
  • 66,625
  • 6
  • 117
  • 269
  • 12
    Excellent answer... but "accountants doing math with big numbers would have alternate systems "? Alternate systems? You can't tantalize us with that and walk away. Any ideas? – Mitch Jun 28 '23 at 20:44
  • 5
    @Mitch Unfortunately I don't know the full details of that, though I'll look into it! – Draconis Jun 28 '23 at 20:49
  • 8
    @Mitch: Could be a nice follow-up question! – Cerberus Jun 28 '23 at 20:49
  • 2
    How did they write about army numbers in big battles? They had extensive records of wars and battles, many of which included tens or even hundreds of thousands of soldiers. The Punic war involved about half a million troops. How did they record that number? – vsz Jun 29 '23 at 04:10
  • 2
    @vsz One immediate possibility but an absolute guess is they wrote them the same way you just did. If the Latin translator I just used is accurate, they could have written quingenta for "half a million". Even if the translator is not accurate, they did have a naming convention for numbers so they could have always spelled things out in words. – Todd Wilcox Jun 29 '23 at 06:04
  • 9
    @vsz: The Roman army is a clear example where they use strict unit size (first hit on Google). By having such a clear unit system and scoping the general logistics to those unit sizes, it means that the numbers themselves are kept low. Half a million soldiers would've been expressed as 100 legions. It not quite the metric system but the factor sizes between units are 8, 10, 6, 10 respectively; which is pretty close and effectively means that you never have to count more than 10 of anything (as you'd then move up to the bigger unit) – Flater Jun 29 '23 at 06:28
  • 1
    The church sign says 1580 - correct ? – Criggie Jun 29 '23 at 23:19
  • 3
    Or it says 1630. – Ray Butterworth Jun 30 '23 at 00:15
  • 2
    @vsz Larger numbers in prose are usually expressed by words or Roman numerals or a mixture (I suspect we can't know for sure which it was originally, as this is the sort of thing that might easily change during manuscript tradition). For thousands of people, the word mille was used, e. g. 20,000 would be milia XX. So for example, in Caesar's Bellum Gallicum (1, 29), I have one version that reads, summa erat capitum Helvetiorum milia ducenta sexaginta tria, Tulingorum milia XXXVI, Latobrigorum XIII, Rauracorum XXIII, Boiorum XXXII, while another version writes the number of Helvetians ... – Sebastian Koppehel Jul 01 '23 at 00:25
  • 2
    ... in numerals too, saying Helvetiorum milium CCLXIII (263,000 Helvetians). Another occasion to write about large numbers were distances, as Romans usually based distance measurement on paces (passus), which is something like 1.5 metres, so there was ample occasion to talk about so-and-so many milia passuum (thousands of paces), from which the word "mile" is ultimately derived. – Sebastian Koppehel Jul 01 '23 at 00:33
  • 1
    Followup question about 'other' numeral systems in Latin – Mitch Jul 03 '23 at 22:23
  • For what it's worth, "year numbers past 1000 weren't a concern" isn't actually true. Most Romans didn't use the Dionysian era, sure, but (a) there were plenty of Jews running around and some were antiquarians or interested in the Anno Mundi, (b) there were Greek philosophers who thought about this stuff as a hobby, and (c) the AUC era -- although never so common as we're usually taught when it's taught -- hit 1000 in the late Classical period. – lly Jul 11 '23 at 15:12
  • @lly True, but the Jews and Greeks had other numbering systems. My understanding is that the vinculum was much more common in Greek than Roman numerals for example. – Draconis Jul 11 '23 at 16:57
  • I know it's a 'me' problem but randos are now downvoting my answer here presumably based on someone misunderstanding what you've posted here. I'd be nice if you stopped by and settled Mitch's continuing confusion one way or the other. – lly Jul 14 '23 at 04:35