As the last year of the last century began there was considerable controversy in the press both in the USA and
in the UK about the correct representation of 1999 in Roman numerals (In the USA see
*Baltimore Sun* 27/12/98; *Washington Post* 31/12/98; in London,
*The Times*, *The Guardian*, and *BBC Online* all dated 1/1/99).
In fact, although there are many alternative ways of depicting 1999 using Roman
numerals, only two or three stand up to scrutiny.

There are several rules used in depicting numbers using the roman numerals I (1), V (5), X (10), L (50), C (100), D (500), M (1000). Some of these were more strictly adhered to than others. Normally, the numerals were simply written out in descending order in a long line so CCXXXV is 235. But another rule allowed the I, X, or C to be placed to the left of a bigger number and subtracted from it. So IV is 4, XIX is 19. The Romans used the subtraction rule, but not always. Doorway numbers at the Colosseum in Rome (c.80AD) show 40 as XL but 44 as XLIIII rather than XLIV.

But one rule is never broken. The Romans strictly represented units, tens, hundreds, and thousands as separate items in their numbers. That is probably because the numerals represented numbers as they were depicted on an abacus - a calculating machine using pebbles or beads which were arranged from right to left in columns of units, tens, hundreds, thousands etc. That means that 99 could be represented as XCIX - 90+9 but never as IC. Similarly, 999 cannot be IM and 1999 cannot be MIM. A consequence of this strict place rule is that an I can only be used to the left of a V or an X; an X can only be used to the left of an L or a C. And a C can only be used to the left of a D or an M.

So the only possible Roman numerical combinations for 1999 are the
following

M (CM or DCCCC) (XC or LXXXX) (IX or VIIII)

1000----900------------------90------------------9

In theory that allows eight different ways of depicting 1999

MCMXCIX

MCMXCVIIII

MCMLXXXXIX

MCMLXXXXVIIII

MDCCCCXCIX

MDCCCCXCVIIII

MDCCCCLXXXXIX

MDCCCCLXXXXVIIII

However, in the Roman examples of Roman numbers which I have seen, where the subtraction rule was used for part of a number but not all of it, then it is the smaller end where it is not used. So you get XLIIII but not XXXXIV. So I would rule out the four examples above which break that rule, leaving as possibilities

- MCMXCIX

- MCMXCVIIII

- MCMLXXXXVIIII

- MDCCCCLXXXXVIIII

Some scholars say that the second is the more accurate, strictly Roman, depiction because the number 9 was usually written VIIII rather than IX. That was certainly true on the Colosseum at Rome where doorway 29 is marked XXVIIII. Others maintain that the fourth, longest version is the purest and would have been most widely understood - you could interpret it simply by counting. The third example may have no genuine Roman validity at all.

However, between Roman times and the medieval period, the principles of
writing numbers in Roman numerals were codified and the subtractive principle
was always used. So today - and for hundreds of years - educated people would
use the most concise form to obey the essential rules - MCMXCIX. And that is what you should see on
monuments and copyright notices to depict the year 1999.

**1999 Question version 1.5
22 July 2005
**